Philip Larson's Blog

One man's opinion...

CONSUMER PRIVACY, NATIONAL SECURITY AND THE GLOBAL FIRM

June 5th, 2007

CONSUMER PRIVACY,NATIONAL SECURITY AND THE GLOBAL FIRM

Philip Larson, 3E, GMUSL, The commercial distribution ofconsumers’ personal information is rapidly increasing. While this informationcan be used to benefit consumers by providing them with more choices and personalizedservices, it can also be misused in ways that invade consumers’ privacy, increasethe threat to national security and inhibit economic growth. Two emergingtechnologies – Service-Oriented Architectures (SOAs) and Business ProcessManagement Suites (BPM) – are breaking down organizational processes intoindividual tasks and other manageable segments. These technologies have madeit more common for an organization’s business processes to consist of a networkof decentralized services. To benefit from cheap foreign labor markets, organizationsare outsourcing more services in their core business processes. While outsourcinghelps organizations reduce costs, it also places consumers’ personalinformation into the hands of a growing number of foreign commercial entitiesincreasing the likelihood of misuse.

To address these privacy andnational security concerns, the United States should enact comprehensivefederal privacy legislation. Part II of this article describes how SOA and BPMtechnology are enabling companies to outsource business processes and therebydistribute consumers’ personal data to growing numbers of global serviceproviders. Part III explains how self-regulation has provided insufficientprotection for consumers’ privacy. Part IV discusses the limitations of theexisting “sectional” approach to United States federal privacy law in regulatingabuses by third-party outsourcing service providers. Part V argues thatCongress should address these deficiencies by enacting comprehensive federalprivacy legislation.

II. THE ADOPTION OF EMERGING TECHNOLOGIES AND INCREASED DEPENDENCEON OUTSOURCING ARE MAGNIFYING THREATS TO NATIONAL SECURITY POSED BY MISUSE OFCONSUMER DATA.

The world is getting smaller. Organizations are adopting emerging technologies that are making it easier totransact with service providers located anywhere in the world. As a result, consumers’personal information is increasingly being distributed to a wider network ofcompanies magnifying the risk of abuse. These data privacy and security abusesare a threat to national security.

A. Business Process Outsourcing (BPO) and Networks ofGlobal Service Providers.

Outsourcing is the practice ofshifting an organization’s operations to a third party vendor.^[1] Business process outsourcing(BPO) occurs when an organization leverages third party services to streamlineany number of processes from administrative support to product development.^[2] The market for BPOis growing rapidly in multiple industries. On a global scale, IDC has estimatedthat the worldwide market for BPO will reach $641.2 billion by 2009, up from$382.5 billion in 2004.^[3] Even Gartner Group’s more conservative estimate for BPO growth suggests themarket in North America alone will be $110 billion by 2009.[4] Moreover, Gartner argues that the BPO market will continue to outpace other ITservices and will see compound annual growth rates of 8.8% through 2009.[5]While there are many models of outsourcing, the two most common are on-shoring- outsourcing to a vendor located domestically - and off-shoring - outsourcingto a vendor in a different country.^[6] While India is the preferred destination for offshore BPO, it is expected therewill be continued growth of outsourcing to China and Eastern Europe.

Organizations have cited manydrivers for this trend. Cost savings is the most obvious and frequently citeddriver, with some estimates arguing that outsourcing can cut costs by 25-30%and up to 50% when off-shored.^[7] For example, a study by University of California at Berkeley found thatprogramming jobs paying $60-80k in the United States go for as little as$8,952/year in China, $5,880 in India and $5,000 in Russia.[8] Outsourcing enables organizations to focus attention on their core competencieswithout the distraction of having to manage non-core services.^[9] Moreover, by using off-shorevendors in different time zones, organizations benefit from consistent,round-the-clock access to these outsourced services. In customer serviceprocesses, this can reduce the difficulty of managing 24/7 customer supportagreements. In product development processes, this can reduce the timerequired to bring a new product to market.^[10]

While the benefits driving adoptionof BPO are clear, there are certain important risks to consider. Companiesfear losing control over core business operations and processes as well aslosing expertise and industry knowledge to these third party service providers.^[11] The financialinstability of outsourcing vendors is another concern. Moreover, as more thirdparties participate in an organization’s business processes data nationalsecurity is threatened from an increased likelihood of privacy and informationsecurity abuse.^[12]

B. Service-Oriented Architectures (SOAs) are EnablingBusinesses to Leverage Services from Global Third-Party Service Providers.

Service-Oriented Architecture (SOA)projects are becoming more common across a number of different industries.^[13] An SOAis an approach to software architecture that exposes an organization’s businesscomponents as reusable “services.”^[14] These services are self-contained, reusable software components that canbe invoked in a standard way by other people and systems over the Internet.^[15]

The scope of a service in an SOAhas a wide range of uses. It may be a simple, one-step task, such as updatingan employee’s home address, or a more complex task involving processing aninvoice or approving a loan application. In the travel industry, for instance,there are services that check hotel availability, book airline tickets, andmake dinner reservations. Each of these autonomous services might be providedby separate vendors and combined by a single organization to create an overall“vacation” process.

An SOA enables applications toeasily pass data over the Internet to invoke services from anywhere in theworld. Therefore, in addition to enabling geographic independence, an SOAmakes it easier for an organization to outsource services in its businessprocesses to third parties.^[16] It is therefore understandable that the adoption of SOA is gaining momentum,particularly in global organizations looking to outsource aspects of itsoperations.^[17] While older applications typically reside in a single geographic location, SOAsare enabling applications to be built as a composition of services provided bymultiple vendors located anywhere in the world. As long as the performance,reliability and security of the services are sufficiently high, they can belinked together as parts of these composite applications.^[18]

Naturally, the implementation anduse of an SOA creates data security and data privacy concerns.^[19] Many technologists and thoughtleaders are currently of the mind that as long as quality of service aresufficient, including performance, reliability and security, it does not matterwhere on the planet the service is provided.[20] The messages exchanged between these services often contain user credentials andother personal information necessary to invoke the service.^[21] This personal information mayinclude names, addresses, Social Security numbers or even credit card andbanking information. As a result, U.S. consumer data is being transferred togrowing numbers of service providers located around the world making it morelikely that the security and privacy of the data will be compromised.^[22]

C. Adoption of Business Process Management (BPM) Softwareis Also Fueling the Growth in Business Process Outsourcing.

Business Process Management (BPM)refers to software used to design, execute, monitor and optimize anorganization’s business processes.^[23] BPM is rapidly becoming the preferred architecture for building agile compositeapplications by linking together services exposed through an organization’sSOA.^[24] According to Gartner, adoption of BPM is on the rise and will continueto grow at a “high rate” through the end of the decade.^[25] In particular, Gartner estimatesthat BPM new license revenue grew 17.3 percent from 2003 through 2004,amounting to $603.4 million in 2004. Moreover, revenue grew across all 10 ofthe geographic regions and subsegments showing that there is a major, globalmarket for this technology.[26]

BPM and SOA technologies complementeach other well. The more business components a company exposes through theirSOA, the more services BPM has to orchestrate within the enterprise processesit manages and automates. Using analytics capabilities, BPM can also helpbenchmark and monitor the performance of the services executing in the processto ensure they are aligned with performance goals.^[27] Therefore, BPM can help reducethe risk organizations face from outsourcing services to third parties byproviding a standard mechanism for evaluating vendor performance and servicereliability. Moreover, BPM makes it much easier for organizations to swapservices in and out of their enterprise processes helping organizations gainagility and adapt quickly to changing business needs. BPM reduces the cost fororganizations to experiment with different combinations of third party serviceproviders enabling them to identify the most efficient combination for theirbusiness. BPM can then encapsulate best practices and ensure these processesexecute consistently and optimally.^[28]

D. The Growing Adoption of BPM, SOA and BPO HasSignificantly Increased the Threat to National Security Posed by Misuse ofConsumers’ Personal Information.

Globalization has forced a“fundamental transformation from regional economies to a single, integratedglobal economy.”[29]Today’s firm may adopt a “follow-the-sun” model to ease the burden of providing24/7 customer support while outsourcing product development services to Russia and payroll processing services to China. The network of global partners in most companies’supply chain management processes are growing. As part of this transformation,consumers have become more aware of privacy and data security issues and theirpotential impact on economic markets and national security.

SOA and BPM are breaking downorganizational processes into individual tasks and other manageable segmentsmaking it easier to swap new services in and out of end-to-end businessprocesses. To stay competitive, organizations are outsourcing many services intheir business processes in order to benefit from cheap foreign labor markets. It is now much easier to collect, analyze and transmit consumer informationinstantaneously to a wider network of affiliates, service providers andpartners.

While this has increased theefficiency and agility of organizations, it has also raised new data privacyand national security concerns. Foreigncompanies and workers are gaining access to some of the most privateinformation about American consumers. One particular concern raised byprivacy advocates has been the threat of misuse by “data brokers,” or companiesspecializing in the collection and distribution of consumer data. Data brokersmust manage the fine line between using information that will benefit consumersand the increased threat to national security that arises from usinginformation in a way that increases the likelihood of identify theft and otherharms.

According to the Federal TradeCommission, the data broker industry is "large and complex and includescompanies of all sizes." Some of the data brokers collect informationfrom public and private sources and use it to provide their own personalizedservices in the marketplace. Others simply resell this information to others,oftentimes with few restrictions on the terms and conditions of thesetransactions, particularly in emerging foreign markets that represent the heartof the business process outsourcing business. The FTC states that the"amount and scope" of the information these data brokers collectdiffers significantly. While many of these uses benefit consumers, such as"fraud prevention, debt collection, law enforcement, legal compliance,applicant authentication, market research", this also makes thesedatabases attractive targets for identity thefts that may use this informationto harm U.S. financial markets or otherwise threaten national security.

While identity theft is a crimepunishable by law in the U.S., a 2003 FTC survey nevertheless estimated that 10million consumers were victims of identity theft in the twelve months leadingup to the survey.[30] The FTC estimated that this misuse of consumers’ personal information resultedin roughly $48 billion in losses and cost general consumers an additional $5billion in out-of-pocket losses. In addition to this significant financialimpact, increased prevalence of identity theft is making it easier and easierto use consumers’ personal information in ways that threaten national security.

By some estimates, in the last twoyears alone over 93 million Americans have had their personal information“lost, stolen, or otherwise compromised.”[31] Paul Kurtz, the head of the Cyber Security Industry Alliance, made thisassessment noting additionally that information security breaches have erodedpublic confidence and represents a “serious threat to economic growth.” Heargues that it is “time for Congress to act” by creating a comprehensivenational law aimed at preventing further data breaches.

While identity theft is aparticularly common way consumers’ personal information is misused it is by nomeans the only threat to privacy and national security. In some instances,threatening to misuse consumers’ personal information has been an effectivebargaining chip for employees to extract personal benefits from theiremployers. For example, in October of 2003 a highly publicized case of misuseof consumer’s personal information involved a Pakistani transcriber doing basicclerical work for the University of California, San Francisco. Thistranscriber threatened to post patients’ confidential information on theInternet unless she was paid more money.^[32] While her motives appeared to have been entirely pecuniary in nature, thethreat that consumers’ personal information will be used in terroristactivities or in other ways that threaten national security are growing.

There have already been examples ofemployees at foreign outsourcing companies using consumers’ personal data tosteal from and defraud American consumers. In April 2005, employees of anoutsourcing company in Pune, India were arrested for the theft of $300,000 fromfour Citibank customers.^[33] Citibank did not find out about the problem until it was notified ofdiscrepancies by its American customers.[34]Reports have been made of Indian gangs offering to pay employees at outsourcedcall centers for Western consumers’ credit card and bank account information.^[35] This number mayeven understate the true risk given that successful hacking of outsourcedcompany’s databases is very difficult to measure.[36] With over 150,000 American tax returns prepared in India in 2004, many fearthat exploitation of personal data and the threat to American economicinterests will only increase.^[37] A number of senators, including Liz Figueroa of California, have argued infavor of privacy legislation that would prevent “outsourcing without anyprotections for privacy.”[38]

The privacy and intellectualproperty laws in common outsourcing destinations like India, China and Russia, are not strict enough to protect consumers.^[39] For example, the U.S. placed India and China on its “priority watch list” of countries that do not provide adequateprotection to intellectual property.[40]Moreover, while there has been “progress” in China’s efforts to enforceintellectual property rights, the State Department still believes the countryhas “a long way to go” before those protections are considered adequate.[41] Additionally, since these processes can involve companies in many legaljurisdictions, there is question as to who may exercise authority when thereare privacy and information security issues.

III. SELF-REGULATION HAS NOT PROVIDED ADEQUATE PROTECTIONOF CONSUMERS’ PERSONAL INFORMATION.

The United States has traditionallypromoted a combination of market-based self-regulation and targeted, sectionallegislation to attempt to prevent misuse of consumers’ personal data in waysthat might weaken national security. This approach has failed to provideadequate protection against misuse of consumer data by foreign companies.

1. Privacy Policies Are Insufficient to Protect Consumers’ From Misuse of their Personal Data.

In 1998, the Online PrivacyAlliance (OPA) was formed to encourage industry self-regulation of privacy.^[42] The OPA createdprivacy guidelines that encouraged two modes of self-regulation: 1) theadoption of privacy policies and 2) the creation of certification groups.^[43] Privacy policiesarticulate the manner in which a company collects, uses, and protects data, andthe choices they offer consumers to exercise rights in how their personalinformation is used.^[44]With privacy policies, consumers may determine whether and to what extent theywish to make information available to companies.^[45] While American law does notrequire companies to post privacy policies, under Section 5 of the FTC Act theFTC has sued companies for failing to comply with their stated privacypolicies.^[46]

Nevertheless, the adoption ofprivacy policies has not provided adequate protection of consumers’ personaldata. American law does not even require companies to post privacy policieslet alone ensure the policies are drafted in ways that actually protectconsumers’ data from misuse. Moreover, having individual privacy policies foreach website means users interested in protecting their information must readthrough the statements of each website they visit in order to understand howtheir information is being protected. Unfortunately, 70% of people in a recentstudy disagreed that “privacy policies are easy to understand,”[47]and few people make the effort to actually read them.[48]In 2003, for instance, an Annenberg Public Policy Center poll claimed 57% of respondentsbelieved that if a company had a privacy policy, they would not share informationwith other entities.[49]

It is an unreasonable burden torequire consumers’ to read all of these statements particularly since most ofthem state that they may be changed at any time. Moreover, it appears thatmany consumers misinterpret the meaning of privacy statements and are simplylulled into a false sense of confidence. For example, a survey conducted bythe Annenberg Public Policy Center found that 75% of adults that use theInternet incorrectly assumed that having a privacy policy meant that a websitewould not share their information with other websites and companies.[50] People assume that if a site has a link to a “Privacy Policy,” their privacywill be protected.

Privacy policies therefore do notsufficiently protect consumers from the misuse of their personal data by thirdparty service providers.

B. Private Sector Certifications Fail to AdequatelyProtect Consumers’ Personal Information.

The Better Business Bureau (BBB),TRUSTe, and WebTrust have created “seals” certifying various levels of privacyprotection for participating websites.^[51] The seal may only be displayed if the company abides by specific privacyprinciples. While advocates of self-regulation suggest that these sealprograms preclude the need for federal legislation, these seal programs havenot proven effective at protecting consumers’ personal data. First, these sealprograms do not carry the weight of law.^[52] Second, they tend to only apply to data provided through an organization’swebsites.^[53] Third, TRUSTe and BBBOnline have been criticized for being mere corporateapologists rather than defenders of privacy.^[54] Regarding TRUSTe, even people central to the establishment of the seal programhave been disappointed with it.^[55] Esther Dyson, who is credited with playing a central role in the establishmentof the seal program, has stated that TRUSTe’s board "ended up being alittle too corporate, and didn’t have any moral courage."[56]

Therefore, while these programscombined with other forms of self-regulation have provided some benefits, afull solution addressing the national security risks of misuse of consumer datawill not be complete without legislation that brings the full weight of thelaw.

IV. THE UNITED STATES’ PATCHWORK OF FEDERAL PRIVACY LAWHAS TOO MANY HOLES.

In addition to self-regulation, avariety of federal laws and regulations regarding data privacy and informationsecurity have emerged. Unlike the broader European approach to privacy law, U.S. privacy law has been more “sectional.”^[57] The United States’ patchwork of privacy legislation regulates how certain typesof entities may use information. Restrictions on the use of consumers’personal information have been extended one vertical at a time and now includeregulations for health care organizations, financial institutions, and consumerreporting agencies. These laws help to strengthen national security byreducing the likelihood that consumers’ personal information will be misused inways that harm the country. Unfortunately, in many situations current federalprivacy laws provides no protection against foreign companies that choose tomisuse consumers’ personal information in ways that harm national security. Moreover,this patchwork approach is creating problems with harmonizing U.S. law with that of the rest of the world.

A. Protecting Consumer Data in Financial Institutions:The Gramm-Leach-Bliley Act.

In 1999, the Gramm-Leach-BlileyFinancial Modernization Act (“GLBA”) was enacted in order to protect theprivacy of consumer information held by “financial institutions.”^[58] The two primarycomponents of the GLBA that govern the collection, disclosure and protection ofconsumers’ nonpublic personal information are the Financial Privacy Rule andthe Safeguards Rule.

The Financial Privacy Rule givesconsumers more control over how and when financial institutions share theirpersonal information.^[59] First, financial institutions are prohibited from disclosing their customers’account numbers to non-affiliated companies when it comes to telemarketing,direct mail marketing or other marketing through e-mail.^[60] Second, when a financialinstitution passes consumer information to a service provider that serviceprovider may only use the information for limited purposes.^[61] If the consumer had no right toopt-out, the service provider may not sell the information to otherorganizations or use it for marketing.^[62] However, if the service provider receives nonpublic personal information from afinancial institution and the consumer does not opt-out, the service providermay use the information for its own purposes or re-disclose it to a thirdparty.^[63]

The Safeguards Rule requiresfinancial institutions to implement reasonable safeguards to prevent misuse ofclients’ nonpublic personal information.^[64] This rule requires the company to develop, monitor and test their program toensure the security of their client’s information. Moreover, this rulerequires companies to select only appropriate service providers and requirethem by contract to implement the safeguards.^[65]

Therefore, while both the FinancialPrivacy Rule and the Safeguards Rule provide some protection from misuse ofconsumer information by third party service providers, the protection islimited to companies providing services to “financial institutions.” Therefore, the GLBA does not protect against abuse by offshore outsourcingvendors that receive consumer information from other types of organizations andinstitutions.

B. Protecting Consumer Medical Records: The HealthInsurance Portability and Accountability Act.

Enacted by Congress in 1996, the HealthInsurance Portability and Accountability Act (HIPAA) required the establishmentof national standards for electronic healthcare transactions.^[66] The HIPAA Privacy Rule, whichtook effect on April 14, 2003, applies to health plans and any healthcareproviders that transmit health information in electronic form.^[67] In particular, the Privacy Ruleprotects all “individually identifiable health information” held or transmittedby a “covered entity” or one of its business associates.^[68] In addition to requiringcovered entities to take reasonable steps to protect the confidentiality ofcommunications with consumers of health care services, it also states that acovered entity may not use or disclose protected health information unless theindividual authorizes the use in writing.^[69]

Therefore, similar to the Gramm-Leach BlileyAct, HIPAA provides some protection against misuse of personal information bythird party service providers receiving health information from health careproviders. However, HIPAA only applies to “covered entities” which consist ofthose who pay for health care “in the normal course of business.”^[70] This definitionwould not provide protection from misuse by many offshore outsourcing vendorsthat receive information from other types of companies and institutions.

C. Section 5 of the Federal Trade Commission Act.

Under the Federal Trade CommissionAct (“FTCA”), the FTC is empowered to (a) prevent unfair methods ofcompetition, including unfair or deceptive acts in commerce; (b) seek monetaryredress for injured consumers; (c) prescribe trade regulation rules definingpractices that are unfair or deceptive; (d) conduct investigations relating to organizationsengaged in commerce; and (e) make reports and legislative recommendations toCongress.^[71]

Section 5 of the Federal TradeCommission Act (“FTCA”) prohibits “deceptive” business practices.^[72] Deceptive practicesare material representations or omissions that are likely to mislead consumersacting reasonably under the circumstances.^[73] The FTC stated in 1998 that using personal information in violation of a postedprivacy policy constitutes a “deceptive practice” and is actionable under theFTCA. Since 1998, the FTC has been quite successful in bringing suits againstcompanies that fail to comply with their stated privacy policies.^[74]

In addition to prohibitingdeceptive practices, Section 5 also prohibits “unfair” practices.^[75] Unfair practicesare those that are likely to cause consumers substantial injury that is neitherreasonably avoidable by consumers nor offset by countervailing benefits toconsumers or competition.^[76] The FTC has used this authority to successfully bring suits against companieswhose practices, while not in direct violation of their stated privacypolicies, still threaten data security. For example, the FTC sued DSW forhaving insufficient security measures to protect credit card and checkingaccount information and found that this constituted an “unfair” practice.^[77] Similarly, BJ’ssettled with the FTC after claims that its failure to encrypt informationstored on its networks and its failure to change default user id and passwordsled to the breach of thousands of consumers’ credit and debit cards numbers.^[78] Most recently, theFTC levied a groundbreaking $15 million fine after finding that ChoicePoint’slack of adequate security measures resulted in a breach that led to over 800cases of identity theft.^[79]

While the FTCA is different fromGLBA and HIPAA in that it is not limited to industry-specific institutions, theFTC has never used its Section 5 authority to bring suit against a company thatprovided consumers’ personal information to a foreign affiliate that thenabused or misused the information. Therefore, it is unlikely that Section 5 ofthe FTCA, in its current form, significantly reduces threats to nationalsecurity that arise from misuse of American consumers’ personal information byoffshore service providers. The recommendation that the FTC push Congress togrant them the authority to hold offshore service providers legally liable foreither direct misuse or personal information or misuse that can be traced backto inadequate security measures is addressed in the next section.

V. COMPREHENSIVE FEDERAL PRIVACY LEGISLATION IS NEEDED TOREDUCE THE THREAT TO NATIONAL SECURITY FROM MISUSE OF CONSUMERS’ PERSONAL DATA.

There is a growing risk to nationalsecurity as businesses adopt emerging technologies that increase America’s dependence on outsourced services. The solution is comprehensive federal privacylegislation. The general public, as well as a growing consortium of privatesector companies, supports national privacy legislation. Moreover,comprehensive federal legislation could be used to harmonize privacyrequirements with those of the European Union creating a compelling model forthe rest of the world.

A. The General Public Supports National PrivacyLegislation.

Opinion polls suggest that amajority of the American public would support national privacy legislation. Ina June 2001 Gallup poll two thirds of respondents were in favor of new federallegislation that would protect online privacy.^[80] In April 2001, the American Society of Newspaper Editors found that 51% ofrespondents were “very concerned” and 30% were “somewhat concerned” that companieswould violate their personal privacy.^[81] In a 2002 Harris Poll, 63% of respondents considered current law inadequate toprotect their privacy and a majority of consumers stated they did not trustbusinesses to handle their personal information properly.^[82]

In particular, consumers have showninterest in legislation that would restrict a company’s ability to providetheir personal information to third parties. A 1991 Time-CNN Poll stated that93% of respondents believed companies should obtain permission from theindividual before selling personal information to a third party.^[83] A March 2000 HarrisPoll found that 88% of users supported requiring a website to obtain consentbefore sharing personal information with others.^[84] Recent surveys suggest that thistrend for consumers to prefer legislation that would provide additionalprotection to their personal information is growing.

Therefore, the general publicappears to support broad privacy legislation that would give them greatercontrol over how companies use their personal data.

B. There is Growing Support in the Private Sector forComprehensive Federal Privacy Legislation.

Traditionally, the private sectorhas been opposed to broad federal privacy legislation believing that regulationcould inhibit growth and innovation. Nevertheless, support for federal privacylegislation has been growing even in the private sector, particularly amonglarge, global firms. Recently, twelve companies formed the Consumer PrivacyLegislative Forum (“CPLF”), an advocacy group to lobby for greater protectionof private information.^[85] The CPLF includes both high tech companies such as Microsoft, Google and eBayas well as companies that haven’t traditionally had a large online presencesuch as Eastman Kodak Co., Eli Lilly and Co. and Procter & Gamble Co. Thebroad range of industries represented by members of the CPLF suggest that newdata privacy issues are not unique to particular industries and that sectional,targeted federal legislation is therefore inappropriate.

The group believes the “time hascome” for “comprehensive harmonized federal privacy legislation” to create a“uniform but flexible legal framework” for protecting consumers’ personal data.^[86] While the CPLF hasnot yet recommended specific language for the statute, the law would likelyrequire businesses to provide notice to consumers when collecting or usingpersonal information and provide individuals control over how the informationis used.^[87]

The members of the CPLF have givena number of reasons for their position in favor of federal regulation. According to Brad Smith, Microsoft’s General Counsel, the complex patchwork ofvarying national and state laws around data and financial privacy has beenconfusing and contradictory and has made it difficult for Microsoft toestablish consistency in their transactions with consumers.[88] It is often unclear what standard will be required in what area of the countrywhen activity that is legal in one jurisdiction may be illegal in another.[89] Nicole Wong, Google’s associate general counsel, agrees that the "unevenpatchwork" of state privacy laws in the United States has made itdifficult and expensive for companies to comply.^[90] Moreover, many companiessupport national legislation because they believe consumer fear of identitytheft and other information security abuses have chilled commerce.^[91]

Therefore, the current approachtowards privacy law in the United States has become burdensome on the privatesector and a growing number of companies believe the time has come forcomprehensive, federal legislation.

C. Federal Privacy Legislation Will Harmonize U.S. Policy with International Laws.

As companies’ enterprise processescontinue to invoke more and more services from around the world to streamlineoperations and implement corporate strategy, consumers’ personal data will passbetween many countries with a variety of different legal standards. Therefore,foreign privacy laws will apply to a growing number of commercialtransactions. It is important for any American legislation to consider theseforeign privacy laws in developing its own privacy legislation in order toprevent conflicting obligations on global businesses.

A comprehensive, harmonized federalapproach to American privacy legislation would be more in line with most of theworld than the existing patchwork approach. The European Union Data ProtectionDirective, in effect since October 1998, created a set of common rules forprotecting personal data and preventing abuse in the EU.^[92] The Directive requirescompanies to ensure that data is collected only for specific purposes, isaccurate and current, and is discarded when no longer needed.^[93] The Directive creates certainobligations on the “processors” of personal data defining the circumstances bywhich the data may be transferred to a third party.^[94] Article 25 prohibits thetransfer of personal information regarding EU citizens to countries that lack“adequate” privacy laws.^[95] Since most countries do not have data privacy laws that satisfy the EUstandards, third party service providers are susceptible to legal challengesunder the Directive.^[96]

For example, the EU determined thatUS privacy laws were inadequate in January 1999. However, the U.S. CommerceDepartment negotiated a Safe Harbor agreement by which U.S. companies can exempt themselves from the Directive. The Safe Harbor requires these companiesto voluntarily adhere to a set of privacy principles including notice, choice,onward transfer, security, data integrity, and access. Other countries thathave not negotiated a Safe Harbor agreement with the EU are liable for securitybreaches that result from their inadequate protections.

The EU Privacy Directive hasdramatically influenced the adoption of privacy law in non-EU countries and isarguably becoming the standard for the rest of the world. Argentina, Australia, Canada, Hong Kong, Hungary, New Zealand and Switzerland have all adopted dataprotection laws that are substantively very similar to the EU. In May 2003, Japan enacted a broad privacy bill applying to any business that uses personal informationdatabases.^[97] While the bill does not specifically declare its jurisdictional reach, itslanguage suggests it will apply to businesses outside of Japan.[98] Additionally, even officials in India have stated that they believe the EUPrivacy Directive is comprehensive and that Indian legislation will be “more orless based on the EU model.”^[99]

Given that a large part of theworld appears to be following the EU model by adopting broad privacylegislation, American legislation must not fall behind or create conflictingrequirements on global businesses. By adopting comprehensive federallegislation, the U.S. can harmonize its privacy requirements with those of theEU and create a unified model for the rest of the world. This will reduce thenumber of conflicting privacy regulations imposed on global businesses. Moreover, federal legislation will create appropriate incentives for theincreasing number of third party service providers to put appropriate securitymeasures in place to protect against misuse of consumers’ personal informationthat could result in threats to national security.

D. The FTC Supports More Comprehensive PrivacyLegislation.

In addition to the general public and agrowing portion of the private sector, the FTC is also in favor of broaderfederal privacy legislation. The FTC has recognized that the protection ofdata privacy and security “is increasingly international in nature.”^[100] They haveeven noted that the globalization of the marketplace means “an increasingamount of U.S. consumer information may be accessed illegally by third partiesoutside the United States or located in offshore databases.”^[101]

Given these structural changes, theFTC has recommended that Congress create a broader, uniform privacy paradigm. For example, the FTC has recommended that Congress extend the “Safeguards Rule”of the GLBA to companies that are not financial institutions.^[102] Currently, the SafeguardsRule applies only to “customer information” collected by “financialinstitutions” and therefore does not cover most data provided to serviceproviders.^[103] Therefore, while the GLBA restricts disclosure of a consumer’s socialsecurity number and address by a financial institution, that same informationis often readily available for purchase on the Internet from a non-financialinstitution.

The FTC should also request thatCongress extend its Section 5 authority to bring suit against companies thatprovide consumers’ personal information to foreign affiliates that do not haveadequate security protections in place. This could simply be an extension of theFTCA’s existing prohibition on “unfair” business practices. Providing consumers’personal information to third party providers that do not have adequatesecurity protections in place is “likely to cause consumers substantial injurythat is neither reasonably avoidable by consumers nor offset by countervailingbenefits to consumers or competition.” The original company collecting theconsumer’s personal information is the natural place for consumers to seekredress should there be any privacy abuses that derive from that initialtransaction. As such, the extension of the FTCA to cover this situation isentirely logical.

VI. CONCLUSION

The adoption of emergingtechnologies such as Service Oriented Architecture (SOA) and Business ProcessManagement (BPM) is helping to fuel growth in business process outsourcing. This is creating a structural change in organizations’ in which businessprocesses are increasingly composed of services provided by geographicallydispersed affiliate and partner organizations. Foreign companies and workersare gaining access to private personal information about American consumerswithout adequate protections in place to prevent misuse. This potential formisuse creates a rising threat to national and economic security. Abuse ofconsumer information by foreign entities is on the rise, decreasing consumerconfidence and inhibiting economic growth. While privacy policies and privatesector certifications have afforded some protection, self-regulation itself hasnot been adequate. Moreover, the United States patchwork of federal privacylaw applies only to specific areas like finance and healthcare leaving too manygaps to be filled with confusing and oftentimes conflicting state laws.

The time has come for comprehensiveprivacy legislation to help address these national security and data privacyissues. The general public and a growing number of companies in the privatesector have recognized this need. Comprehensive legislation would help the U.S. harmonize its privacy policies with the international community protecting globalcompanies from the threat of conflicting legislation. Moreover, the FTC hasacknowledged that broader legislation extending the FTCA would enable it tomore effectively protect the privacy interests of consumers against misuse andabuse by third party service providers.

Posted in Law and Technology | No Comments »

DOES GOOGLE BOOK SEARCH VIOLATE COPYRIGHT LAW?

June 5th, 2007

DOES GOOGLE BOOK SEARCH VIOLATE COPYRIGHT LAW?

A LEGAL AND POLICY ANALYSIS

Philip Larson, 3E, GMUSL, plarson2@gmu.edu

 

I. INTRODUCTION

The debate over the appropriate role of copyright law in the new digital age is a contentious one. At the heart of this debate is the question of how copyright law should govern the tradeoff between encouraging the creation of innovative works and maximizing secondary uses derived from those works. By providing creators with certain exclusive property rights over their expressive works, copyright law creates financial incentives to invent and create. Without these incentives, the argument goes, fewer people would spend their time and effort generating valuable intellectual property. On the other hand, copyright law must not provide excessive barriers for secondary users to build on existing ideas and innovations to create new expressive works. New advancements tend to build on past innovations. Copyright law must therefore attempt to facilitate widespread distribution of expressive works so that they may be used and built upon. This balance between creation and use is a delicate one. Too much protection for creators will lead to suboptimal use of these ideas in society and thereby raise the cost of future innovation. Too little protection for creators and fewer innovations will be created in the first place.

The digital revolution is creating new challenges in identifying the appropriate balance between creation and use. We live in an age in which digital copies of creative works can be copied expeditiously and with little effort. Copyright law must place limits on the rights of individuals to make these copies in order to protect the interests of the initial creators. However, the scope of these limits cannot be so broad as to stifle future innovation.

Google’s new Book Search initiative is a timely example of the challenges that arise in identifying the appropriate scope of copyright. As part of the initiative, Google has begun scanning collections of several prominent libraries in order to create a vast searchable database of literary works. Copyright holders who have not authorized this scanning have filed suit claiming the creation of these digital copies constitutes direct infringement of their exclusive rights to copy and distribute these works. Google claims its actions are protected under the doctrine of fair use. This paper discusses the legal and equitable merits of these conflicting claims.

Determining whether Google Book Search violates copyright law is not clear cut. Given the unique facts and issues presented by Google Book Search, there is very little legal precedent to definitively dispute or legitimize each party’s claims. Part II of this paper provides the reader with background on the Google Book Search initiative, and in particular the more contentious Library Program. Part III of this paper provides a brief overview of the relevant copyright law including the doctrine of fair use. Part IV provides a legal analysis of whether Google Book Search qualifies for protection from copyright infringement under the doctrine of fair use. This paper suggests that the current state of copyright law makes it impossible to predict who will triumph based purely on legal precedent. Part V sets forth economic considerations that tend to favor a finding of fair use. Part VI concludes that while precedent provides little guidance on the appropriate legal outcome in this case, economic analysis leans in favor of finding that Google Book Search is protected under fair use.

II. WHAT IS GOOGLE BOOK SEARCH?

In December 2004, Google announced plans to digitally scan books from prominent libraries as part of their Google Book Search project. [1] The Google Book Search project consists of two programs: 1) the Partner Program, and 2) the Library Project. [2] In the Partner Program, copyright owners enter into an agreement with Google authorizing Google to scan the full text of their copyrighted works. When a user enters a query into the Google search engine, the engine scans its indexes and provides the user with a list of bibliographic information on books that match the search terms. [3] Included in the search results is content from the book that surrounds the text matching the user’s query. [4] The amount of content available to the user depends on the agreement Google has with the copyright holder. The Partner Program has not raised potent copyright concerns because the program is conducted with the approval and consent of the copyright holders. By contrast, Google’s Library Project has raised multiple copyright issues and is the subject of two current lawsuits. [5]

The Library Project aims to provide “an enhanced card catalog of the world’s books” by making the collections of major libraries searchable and discoverable through the Google search engine. [6] Without approval from copyright holders, Google has begun scanning over 25 million books from major libraries around the world including those of Princeton University, Harvard University, Stanford University, and Oxford University. [7] In exchange for providing the libraries with a digital copy of the materials for their own use, Google retains an indexed digital copy of the work for use within its search engine. [8] When a user submits a query, Google Book Search displays results to the user from its scanned book repository. The amount of free text made available by Google depends on the status of the copyright on the work. For books that are in the public domain, Google enables the user to browse the entire contents of the work. For books that are still in copyright, Google restricts the information to end users proving only small “snippets” of text surrounding the search terms. [9] Moreover, Google has instituted an “opt out” policy for the Library Program. Using the “opt out” option, publishers provide Google a list of books that they do not want included in Google’s online index. [10] In this way, copyright holders may prevent their works from being made available through Google’s Book Search technology.

The major copyright issues with the Google Library Project arise from Google’s decision to scan books from the libraries that are still in copyright without seeking the express consent of the copyright holder. The primary legal issue is whether Google’s Library Project violates copyright law or whether Google’s actions are protected under the doctrine of fair use.

III. COPYRIGHT LAW AND THE DOCTRINE OF FAIR USE

Copyright law confers on copyright holders certain exclusive rights over their works. Direct copyright infringement occurs when those exclusive rights are infringed by a third party. Under certain circumstances, the fair use doctrine provides a defense against a copyright infringement claim. In order to assess whether Google Book Search violates copyright law, this section provides a brief overview of applicable law as well as background on the fair use doctrine.

A. Copyright Law Provides a Set of Exclusive Rights Over Creative Works.

While the term “copyright” does not appear in the Constitution, Article I gives Congress the power “[t]o promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries.” [11] Congress first exercised this power when it passed the Copyright Act of 1790 and copyright law has been updated significantly since then. [12]

Copyrights are the set of exclusive rights granted by the government to generators of creative or artistic work. [13] Copyright law confers on the creators of expressive works the right to 1) produce copies of the work, 2) display the work publicly, and 3) distribute copies of the work. [14] Copyright is sometimes considered a “negative right” in that it prevents third parties from taking certain actions with respect to the copyrighted work. [15]

B. Fair Use: An Affirmative Defense to Copyright Infringement.

Copyright law in the United States does not prohibit all unauthorized uses of copyrighted materials. The Copyright Act of 1976 established a “fair use” doctrine that allows certain uses of copyrighted material even without the permission of the copyright holder. [16] The fair use doctrine is an affirmative defense which need only be invoked once the plaintiff shows “prima facie” evidence of copyright infringement. [17] At that point in litigation, the defendant then bears the burden of proving his use was fair and therefore does not constitute copyright infringement.

The fair use doctrine provides a framework for courts to assess whether a particular use of a work constitutes copyright infringement. Rather than list specific permissible uses for each type of work, the Copyright Act lists out four factors the court must take into consideration when determining whether the fair use doctrine applies. [18] These factors include 1) the purpose and character of the use, including whether the use is of a commercial nature or is for nonprofit educational purposes, 2) the nature of the copyrighted work, 3) the amount and substantiality of the portion used in relation to the copyrighted work as a whole, and 4) the effect of the use upon the potential market for or value of the copyrighted work. [19] Nevertheless, f air use must be applied in light of the “overall purposes of copyright law” which allows other case specific factors to be taken into consideration. [20]

Courts use the first factor, the “purpose and character” of the use, to identify whether the use helps stimulate creativity for the benefit of the general public or whether it merely supersedes the copyrighted work for the personal benefit of the infringer. Courts look at the extent to which the use is “transformative” rather than merely “derivative.” [21] Uses that “transform” the original work by infusing it with new meaning or new understanding are more likely to be protected under fair use. [22]

Regarding the second factor, the “nature” of the copied work, courts assess the usefulness to society of having the work in the public domain. Courts provide greater protection for creative works than for factual works whose contents more rightly belong in the public domain. Moreover, strong “public interest” in allowing the use can help defeat a copyright infringement claim. [23]

The third factor assesses the “amount and substantiality” of the original copyrighted work that has been used in the new work. Typically, the less of the copyrighted work the defendant uses the more likely it will be considered fair use. However, in some cases courts have upheld the fair use defense even when the entire work has been copied without the authorization of the copyright holder. For example, in Sony the Supreme Court held that “time shifting”, or taping television shows for viewing at a later time, constitutes a fair use of copyrighted programming. [24] Additionally, in a particularly relevant case in the Ninth Circuit, the court held that a defendant’s copying of entire copyrighted photographs in order to create thumbnail images for its search engine results was fair use. [25]

Under the fourth factor, the “effect of the use on the potential market” for the copyrighted work, courts assess the extent to which the copyright owner’s ability to exploit his original work has been hurt. The courts evaluate the impact of the defendant’s use on both current and potential markets for the original work. The Supreme Court has called this factor the “single most important element of fair use.” [26] However, it has also stated that all four factors are to be explored, and the results “weighed together, in light of the purposes of copyright.” [27]

IV. DOES GOOGLE BOOK SEARCH VIOLATE COPYRIGHT LAW?

Applying current copyright law to the Google Book Search debate leads to ambiguous results. The unique facts and issues presented by Google Book Search and the dearth of relevant legal precedent makes it nearly impossible to predict who will triumph. This section identifies which of Google’s activities constitute prima facie evidence of direct copyright infringement and then analyzes the strength of Google’s fair use defense.

A. Google’s Unauthorized Scanning of Copyrighted Works Constitutes Prima Facie Evidence of Copyright Infringement.

To prove Google Book Search violates copyright law, the copyright holders must first provide prima facie evidence of direct copyright infringement. The copyright holders will have no trouble meeting this burden. Google creates an unauthorized copy of a protected work each time it scans copyrighted books from its partner libraries without the permission of the copyright holder. This clearly violates the copyright holders’ exclusive right to produce copies of the work under the Copyright Act. [28] Therefore, unless this scanned copy is protected by an applicable defense, it constitutes a violation of copyright law.

Two other actions by Google might constitute prima facie evidence of copyright infringement. First, Google displays “snippets” of copyrighted text in search results. However, since the amount of copied text is small, these “snippets” are likely to be treated like quotations which generally are not violations of copyright. Second, Google creates digital copies of the copyrighted books for the libraries. However, the current lawsuits against Google Book Search do not claim infringement based on these library copies. Therefore, this paper will only address the copyright issues associated only with the scanned copy Google makes for inclusion in its search engine.

B. Google’s Opt-Out Policy Does Not Provide a Defense to Copyright Infringement.

As described above, Google’s opt-out policy offers copyright owners the ability to provide Google with a list of books that they do not want included in the program. This opt-out policy does not provide Google with a valid defense to direct copyright infringement. Copyright law provides copyright holders the exclusive right to copy and distribute their works. As such, copyright law defines an opt-in paradigm in which third parties that would like to use a protected work must first receive the permission of the copyright holder. The primary exception to this opt-in paradigm is the fair use doctrine. The nature of these exclusive rights in protected works puts the burden on Google to request permission to scan the work unless that scanning constitutes fair use.

In addition to legal precedent, there are strong practical reasons why opt-out policies are not valid defenses to copyright infringement. An opt-out paradigm would quickly become overly burdensome on copyright holders. While opting out of Google Book Search may not be overly burdensome, a general opt-out paradigm would require copyright holders to opt out of all organizations’ digitization initiatives. [29] Copyright owners would be at an informational disadvantage, constantly threatened by the emergence of new digitization initiatives of which they might not be aware. Under an opt-out paradigm, specialized indices from niche vendors would likely emerge. Microsoft and Yahoo have already announced large-scale digitization efforts. An opt-out paradigm will quickly become unmanageable for copyright holders.

Therefore, legal and practical reasons prohibit Google’s opt-out policy from providing a valid defense to copyright infringement.

C. Analysis of the Four Fair Use Factors Provides No Clear Answer as to Whether Google Book Search Constitutes Fair Use.

Given that there is prima facie evidence that Google Book Search infringes copyrights and its opt-out policy provides no defense, Google will only prevail if its actions constitute fair use. The primary legal question, therefore, is whether Google’s creation of unauthorized digital indexed copies of copyrighted books for inclusion in a search engine database is protected under the fair use doctrine. If it is protected, Google escapes liability. If it is not protected, Google Book Search likely violates copyright law. This section applies the fair use factors to the facts in this case and concludes that the factors do not weigh heavily in either parties favor.

1. The “Purpose and Character” of the Use Weighs Slightly Against Fair Use.

Google’s unauthorized use of the copyrighted works is for a commercial purpose. Google is a corporation whose goal is to maximize shareholder wealth and presumably intends for the Book Search service to provide additional sources of advertising revenue. The commercial nature of Google’s initiative weighs in favor of the copyright owners. On the other hand, Google is not attempting to profit from the direct sale of copies of the books that it has scanned into its database and is thus not highly exploitive of the protected work. [30] Selling advertisements based on user search queries is arguably not competitive with the direct sale of the books.

Moreover, Google’s new use is transformative. Google augments the original unauthorized copy with indexing information and metadata such that the new copy bears very little resemblance to the original. The goal of Google Book Search is not to replace the market for the original books but is rather to make it easier for users to discover books of interest to them. [31] Most users will not use the tool as an alternative to buying the book itself because only small snippets of the original work will be provided. In fact, publishers and authors tend to agree w/ Google that the Book Search initiative will most likely increase book sales.

Therefore, while the “purpose and character” of Google’s use is commercial in nature and weighs against fair use, the transformative nature of the use suggests that the court is unlikely to find a strong inference based on this factor alone.

2. The “Nature” of the Copyrighted Work Weighs Slightly in Favor of Fair Use.

The second factor of the fair use doctrine is the “nature” of the copyrighted work. The Google Book Search initiative will encompass large numbers of both factual and creative works, the latter being entitled to the greatest levels of copyright protection. While the “majority” of the works scanned under the project will be non-fiction “fact-based works,” this distinction will probably not matter to the court given the large total number of works in each category. [32] Moreover, all works that have been published are generally afforded more protection by the fair use doctrine because the expression of the artist’s work “has already occurred.” [33] Therefore, at first blush it would seem that this factor would weigh against fair use.

However, courts looking at the “nature” of the work also consider the “public interest” in allowing the new use. [34] Google has a strong argument that public interest in the Google Book Search program is large and should outweigh the fact that many of the works being copied are creative published works. Google Book Search provides a new, innovative way to research and discover literature. Courts have frequently found that usefulness to the public is a critical issue in analyzing this particular fair use factor. The benefits to society from Google Book Search clearly favor of fair use.

Therefore, while Google’s use involves copying large numbers of creative published works, the court will likely balance this against the considerable benefit the service provides to the public. While somewhat uncertain, this factor will probably weigh slightly in favor of fair use.

3. The “Amount and Substantiality” Factor Weighs Slightly Against Fair Use.

The third fair use factor is the “amount and substantiality” of the copyrighted work being used. Google makes a full copy of each work and stores the entire digital copy, along with metadata and indexing information, in its search database. The court will consider this copy to be a “substantial portion” of the work as a whole. This weighs heavily against fair use. On the other hand, there is legal precedent holding that this factor should not be weighed as heavily in situations where the defendant copies “only as much as is necessary for his or her intended use.” [35] The only way Google can provide its Book Search service is if it stores a full, indexed digital copy of the original work in its search database. Therefore, Google is only using as much of the work as is necessary to provide the service.

Additionally, this type of “intermediate copying” can constitute fair use even when a full reproduction of the work is needed up front to create the noninfringing end use. For example, case law involving software translations has held that even entire reproductions can be protected by fair use if they are used to create noninfringing end uses to consumers. [36] While Google makes an entire copy of the work up front, they only show a tiny “snippet” to end users. Therefore, it is possible that Google’s full reproductions may be protected as fair use since they are essential to the creation of a noninfringing end use.

The copyright holders will argue, however, that these “intermediate copying” cases arose in order to address a unique problem with software – that full translations were the only means of extracting the underlying unprotected ideas contained in the machine-readable source code. By contrast, intermediate copying is not necessary to extract the unprotected ideas from books.

Therefore, while the amount and substantiality of the work Google is copying is large, Google is only using as much as is necessary to provide a valuable noninfringing service to the public. Despite the public benefits, this factor is likely to weigh slightly against Google’s fair use defense.

4. The Effect of the Use on the Current and Potential Markets for the Work Weighs Slightly In Favor of Fair Use.

The fourth fair use factor is the effect of the use on the current and potential markets for the work. Google Book Search will probably not be considered a substitute for the copyrighted work. The service does not provide end users with the entire contents of the book. It limits end users access to mere “snippets” of content. In most cases, these snippets are unlikely to decrease demand for the book itself. Rather, most publishers and authors agree that the program will probably increase the demand for their copyrighted works. Google Book Search makes it easier for potential buyers to find the books they want. It facilitates purchasing the books by directing buyers to any of a number of online stores. Therefore , Google Book Search will likely serve to aid the public with their book purchasing decisions. Google will argue that its search index is simply a next-generation card catalog or compilation. Courts have routinely dismissed cases in which indexes, compilations and catalogs have been charged with violating fair use. Moreover, Google book search may even create new potential markets for works that are in-copyright but out of print. In these cases, copyright holders have very little current opportunity for economic gain. By combining Google Book Search with emerging “print on-demand” technologies, Google may help facilitate a brand new market for consumers to purchase out-of-print books.

On the other hand, copyright holders will argue that Google Book Search will harm certain markets for their works. While the display of a few sentences of text around the users search criteria probably would not be enough to reduce demand for purchasing the entire book, this practice becomes a slippery slope. Google has an incentive to provide as much of the copyrighted work as possible without infringing copyright. The more content provided in Google’s “snippets”, the less need the user will have to purchase the original.

Additionally , copyright holders will argue that Google Book Search prevents them from licensing their works to internet-based search engine indices like Google. As digitization projects increase, the potential market for licensing content to online providers could grow significantly. Copyright holders will argue that their exclusive rights to copy and distribute their expressive works entitle them to negotiate and share in some of the value Google Book Search creates. In particular, they will argue that Google should pay publishers a license fee for the right to scan and store a digital copy of the copyrighted work for use in their search engine. By refusing to engage in this negotiation, Google is effectively circumventing the copyright holders in an attempt to illicitly derive profits from other people’s intellectual property. Additionally, by providing libraries with a digital copy of the copyrighted content, Google is arguably destroying the publishers’ potential market of selling digital copies directly to libraries. Under this view, Google Book Search has harmed both current and potential markets for the copyrighted works.

Further complicating the issue is the fact that the Google Book Search program exposes copyright holders to new risks. In a digital age in which identity theft and hacking is rampant, who is liable if Google’s database is compromised? Google’s search engine repository holds unauthorized digital reproductions of other people’s intellectual property. Copyright holders are compelled to depend on security measures put in place by Google that they may not consider adequate. Given the ease by which digital works may be copied and distributed over the internet, a security breach of Google’s index could be devastating. A single digital copy propagating over the internet could drastically cannibalize the future market for compromised works. Allowing Google to maintain this database forces these security risks upon copyright holders against their will.

Therefore, it is not clear how a court will assess the validity of these various arguments in determining whether this fair use factor weighs for or against fair use. Google Book Search will likely open new markets for books while impeding others. In some instances sales of works may increase while in others they may decrease. On the whole, however, the court will probably be persuaded that the public benefits of Google Book Search cause this factor to weigh slightly in Google’s favor.

5. Summary of the Four Fair Use Factors: Is the Argument a Tie?

The above analysis of the four fair use factors above suggests that there is no clear answer as to whether Google Book Search violates copyright law. Two of the factors appear to weigh slightly in favor of fair use while the other two factors seem to weigh slightly against fair use. It is likely that the court will have to consider additional factors in determining whether the fair use doctrine provides Google with a valid defense to copyright infringement.

V. AN ECONOMIC ARGUMENT IN FAVOR OF GOOGLE BOOK SEARCH.

While courts tend to focus on the four fair use factors, the Supreme Court has stated that fair use is an “equitable rule of reason” that permits courts to “avoid the rigid application of the copyright statute” when it would stifle creativity. [37] Therefore, other arguments may be considered by the court in makings its judgment of fair use.

Intellectual property rights should be designed to maximize consumer welfare over the long run by incentivizing innovation while also encouraging secondary uses that build on previous innovation. An important question, therefore, is whether the benefits to society of protecting Google Book Search under the doctrine of fair use outweigh the direct and opportunity costs.

There is strong evidence to suggest that the benefits to society from protecting Google Book Search under the doctrine of fair use will be very large. Google’s stated mission is to organize the world’s information and make it universally accessible and useful. [38] This mission encourages future innovation and creative works by getting the right information to the right people when they need it the most. With Google Book Search, users who seek hard to find books that appear in libraries but are difficult to find elsewhere will be better off. Google Book Search will help protect works that are only available in print and may be degrading in quality that might otherwise pass into obscurity as their pages age and fade. Google Book Search will open new markets for out-of-print books, and will likely improve markets for most books that are currently in print. Google Book Search will provide an incredibly valuable tool for students, professors and others to discover books relevant to their research initiatives. Moreover, whole new markets will likely emerge providing value added services on top of Google’s growing repository of information. In short, Google Book Search will provide great value to society by streamlining the process of discovering, identifying and retrieving information that may be used by the next generation of innovators and creative thinkers. Moreover, the costs to society of protecting Google Book Search under fair use are relatively low. A finding of fair use is unlikely to disincentivize the initial creation of creative works. Authors and publishers agree that sales of books are likely to go up by having their works more widely accessible through the Google Book Search program. Therefore, while some harm may result to copyright holders in lost market opportunities for their works, these harms are greatly outweighed by the sum of the benefits to consumers.

By contrast, if the court held that Google Book Search violates copyright law, the benefit to society would be small and the costs would be quite large. If Google was forced to negotiate with each individual publisher or copyright holder in order to get permission to scan in-copyright books, the transaction costs would likely be prohibitive. For example, these transaction costs would include the costs of finding out who owns a particular copyright. Copyright notices provided with books in the U.S. do not specify whether the author, the publisher or a third party owns the copyright. International books often lack copyright notices altogether. Moreover, the U.S. has no federal copyright registry. Therefore, simply identifying the current copyright owner for each in-copyright book would be very costly. In addition to the costs of identifying the owner, the transaction costs would include the cost of locating the owner. In some cases, the owner may have passed away and the copyrights passed on to a distant heir who may or may not be aware that they are the copyright holder. Additionally, since there is no national registry it would be very difficult to identify the current address or contact information for certain copyright holders.

The transaction costs would also include the costs of negotiating a license with each individual copyright holder once the holder has been identified and located. About 75% of in-copyright books are currently out-of-print. In many cases, these copyrights are owned by the publishers. The publishers in these circumstances would have no economic incentive to allow Google to scan the books because the books are out of print and the program would therefore not increase their revenue. The original authors who have an interest in including their works in Google Book Search to reach a wider audience would not have the legal right to authorize Google to scan their works. Therefore, many of these out-of-print books would probably not be included in the index.

When the transaction costs of identifying, locating and negotiating with a copyright holder for a particular work are multiplied by the millions of in-copyright books Google plans to scan, the costs quickly become prohibitive. Therefore, if Google is forced to obtain permission to scan in-copyright works and permission is incredibly difficult or in some cases impossible to obtain, a project like Google Book Search becomes very unlikely to succeed. The costs of voluntary exchange through contract are so high relative to the potential marginal benefit that no exchange would take place. At best, Google Book Search would provide an incomplete search index that would prevent the public from quickly and easily browsing and querying the full universe of relevant books.

Therefore, economic analysis seems to suggest that carving out a fair use exception from copyright law for the Google Book Search initiative would maximize consumer welfare. Google Book Search is likely to provide great benefits to society at very little cost. By contrast, a finding that Google Book Search violates copyright law would make it unlikely that a service like Google Book Search would ever be provided. Consumer welfare will suffer. Therefore, equitable and economic analysis argues in favor of fair use.

VI. CONCLUSION

Intellectual property laws should be written and interpreted to maximize consumer welfare. Interpreting copyright law in the digital age can be challenging because of the difficult tradeoff between incentivizing innovative creation and optimizing the use of those creations in future works. Google Book Search is a great example of a innovative service that raises new issues of how copyright law should be applied in the digital age. A purely legal analysis does not provide a clear answer as to whether Google Book Search violates copyright law. When precedent and statutes provide little guidance, it is appropriate and necessary for courts to look to equitable and economic arguments to help guide their decisions. Here, Google Book Search should be protected under the doctrine of fair use because the alternative would place a crippling tax on innovation and prevent valuable services from being provided to the public. A finding of fair use will maximize consumer welfare and will not substantially reduce the incentives of authors to write and distribute new creative works.

Posted in Law and Technology | No Comments »

DRAFTING THE LLC OPERATING AGREEMENT FOR A SOFTWARE VENDOR

May 4th, 2007

DRAFTING THE LLC OPERATING AGREEMENT FOR A SOFTWARE VENDOR

Philip Larson, 3E, GMUSL, plarson2@gmu.edu

 

I. INTRODUCTION

This paper analyzes issues that may arise in drafting the operating agreement of a Limited Liability Company (“LLC”) formed by ten computer programmers for the purpose of developing and selling software products. The ten programmers will become part owners and members of the LLC. Each programmer has the same approximate level of programming experience and will contribute roughly equal small amounts of capital. After developing an initial set of software products, the owners plan to raise additional capital to bring the products to market and fund future research and development.

For the purposes of this paper, certain assumptions have been made. First, each member is presumed to bring a unique set of technical skills to the LLC that are complementary rather than duplicative. Second, the members are presumed to have good working relationships from which they derive efficiency gains. Therefore, the owners may be concerned about bringing in new members that may disrupt this working dynamic. Third, it is presumed that developing new software products frequently takes a number of years. Therefore, the LLC’s revenue during these start-up years is likely to be low making the LLC’s intellectual property one of its most valuable assets.

II. FINANCIAL RIGHTS AND DUTIES OF MEMBERS

A: Member Contributions and Accounts

LLCs have great flexibility in defining the financial rights and duties of their members. LLC statutes frequently allow capital contribution to be in cash, property or services. [1] Operating agreements often list the amounts, value and time of each member’s initial contribution.

1. Service Contributions Do Not Count Towards Capital Accounts

In small technology start-ups, the owners often do not contribute much capital in the form of cash or property. Each owner’s primary contribution consists of the ongoing technical or business skills and services they provide to the firm. However, service contributions are difficult to valuate. Here, each member has similar technical experience and their service contributions are likely to be comparable. Therefore, to help avoid valuation problems, the operating agreement should specify that only cash and property contributions, not services contributions, are to be included in the valuation of capital contributions and capital accounts. Moreover, the operating agreement should make clear that these contributions are equity investments that earn no interest. And finally, the paragraph 4.3 of the Chameleon Agreement (“CA”) should be modified to provide that members are not entitled to compensation adjustments based on disparate service contributions.

2. Mitigating the Risk of Member Shirking in Service Contributions

The primary contributions of the members of the start-up technology LLC will be their service contributions. If the members are not rewarded according to their individual contribution to the LLC’s profits, each member may try to “free ride” off the other members’ efforts. On the other hand, valuating each member’s contribution to the LLC’s profits is difficult. The administrative costs of implementing a compensation system that perfectly rewarded the members’ contributions would be prohibitive. Therefore, the operating agreement must encourage members to participate fully in the LLC while minimizing administrative costs and the risk of shirking.

To encourage member participation, the operating agreement should include a provision in paragraph 7.8 of the CA requiring each member to devote substantially full-time to the performance and management of the LLC. To discourage shirking, paragraph 10.1 of the CA should specify that excessive shirking or devoting less than full-time to the LLC are grounds for expulsion. This provision impacts the fiduciary duties of the members because portions of the expelled member’s interests will likely revert back to the remaining members, creating an opportunity for self-dealing. Therefore, the conditions for expulsion based on shirking must be clearly defined to reduce the risk of opportunism by the other members.

B. Membership Interests

Each member of an LLC generally holds a “membership interest” that entitles them to share in the profits of the LLC. Unless the operating agreement provides otherwise, profits of the LLC are allocated to the members on the basis of their capital contributions. Given that the capital contributions and technical experience levels of the programmers is about the same, paragraph 4.2 of the CA should be modified to provide that each member is to receive a 10% interest in the LLC regardless of any slight differences in members’ capital contributions or experience levels.

In a small technology firm, the owners should anticipate the need to support new classes of membership interests for future investors (both active and passive), managers, employees, strategic partners and other potential stakeholders. A major benefit of the LLC compared with other business forms is its complete flexibility in structuring classes of membership interests, each with its own financial and management rights. For example, the limited liability characteristic of the LLC form eliminates the problem in Gay Jenson Farms Co. v. Cargill, Inc., in which a creditor became personally liable for the debts of a partnership when he became so involved in management of a partnership that he was held to have “de facto control.” [2] Unlike the partnership in Cargill, the LLC would enable a venture capitalist or creditor to obtain some control over management of the LLC while still insulating them from personal liability. LLCs are also more flexible than corporations, which require certain formalities to be followed and, in the case of S corporations, have shareholder and class of stock restrictions. Therefore, paragraph 9.4 of the CA should be modified to include language that creates a structure for defining and allocating future classes of membership interests. Moreover, the provision should define the procedures (e.g. voting requirements, etc.) needed for their creation and distribution.

III. MANAGEMENT AND CONTROL RIGHTS

An LLC has great flexibility in defining its management and control structure. Members describe how the LLC will be operated, managed and controlled in an operating agreement. One of the major advantages of the LLC is that, unlike a corporation, no formalities must be observed in the management of the firm in order to maintain the limited liability of its members. [3]

A. Management Structure: A Middle Approach

An LLC may typically be organized for “any lawful purpose” and may do all things “necessary and convenient” to carry on its affairs. [4] In most LLC statutes the default rule is that LLCs are to be managed directly by their members. However, LLCs may choose to adopt a more centralized management structure in order to streamline daily decision-making. The flexibility to adopt a varied and unique management and governance structure distinguishes the LLC from more rigid corporations and general partnerships.

Decision-making in a technology LLC with ten members might be significantly frustrated by a default rule requiring all firm decisions to be made by a vote of the members. Therefore, the owners of the technology LLC will probably want to use the flexibility of the LLC and concentrate certain management powers in one or more members. These provisions would modify paragraphs 6.1 and 8.2 of the CA which define the management and control rights. By modifying the default “member-management” structure, the LLC could reduce its potential liability by restraining non-managers ability to bind the firm in inappropriate ways. On the other hand, delegating management power also increases the risk of opportunism. Managers may take self-interested actions at the expense of the non-managing members. Therefore, the operating agreement must balance the need to protect managers’ ability to exercise discretion with the need to prevent managerial abuse of power.

One approach to balancing these competing interests would be for the technology LLC to adopt a middle approach between member-management and manager-management by defining and empowering roles commonly that are common to businesses. For example, the owners might modify paragraphs 6.1 and 8.2 of the CA to define certain key roles, such as the CEO, CFO, CMO, COO and CTO, and empower these roles to make certain decisions without requiring a vote of all members. By mimicking roles common to business in the management structure of the LLC, the owners will make it easier for investors and third parties to determine who has authority in various business transactions. These provisions should also be updated to define the procedures by which these empowered roles are elected and removed. To address the concern of opportunism inherent in this more centralized management structure, these roles should be restrained from taking unilateral action in “extraordinary” matters of the firm. The scope of these extraordinary matters is discussed below.

B. Voting Rights: Supermajority Requirement for “Extraordinary” Decisions

In a manager-managed LLC, each manager generally is entitled to one vote, and a majority vote of managers is ordinarily required to approve most decisions affecting the LLC. [5] In a member-managed LLC, members typically have equal rights in the management and conduct of the company’s business. [6] By taking a middle approach, the technology LLC will enable members in key roles to make certain decisions without requiring a vote of all members while restraining these members from acting unilaterally on “extraordinary” matters. Therefore, paragraph 8.1 of the CA should be updated to require a supermajority vote of members for all “extraordinary” decisions. For clarity, these extraordinary matters should be listed out explicitly in the agreement and should include decisions regarding mergers, acquisitions, asset sales, raising capital, adding new members, and selling or leasing the LLC’s intellectual property. This will protect the LLC against issues like in Patel v. Patel in which a partner sold a major asset of the partnership without the approval of the remaining partners. [7]

C. Restrictions on the Admission of New Members

In a small technology firm, the admission of new members could seriously disrupt the efficiencies of the existing management structure. Additionally, admitting new members with equity interests will dilute the interests of the existing members. Under most statutes, the LLC may admit new members with a mere majority vote. [8] To reduce the risk of unnecessary disruption and dilution, the technology LLC should modify paragraph 9.4 of the CA to require a supermajority vote to admit new members.

D. Public Offerings, Mergers and the Potential Need for Conversion

Technology businesses are good candidates for public finance given their need for accelerated growth once the software product is ready. Dramatic infusions of capital may be necessary to build up the marketing and sales operations needed to effectively bring the product to market. Here, the owners anticipate the need to raise additional capital after they develop their initial set of software products. Unfortunately, the LLC structure can sometimes complicate public offerings and mergers. Investors more comfortable with investing funds in the better understood corporate form may view the LLC with suspicion and lower their valuations. Therefore, the owners may need to convert the LLC into a corporation prior to IPO. Alternatively, the owners could merge the LLC into a shell corporation in anticipation of an IPO. To facilitate these potential conversions or mergers, the operating agreement should include a provision making it clear that mergers and conversions are not prohibited, particularly in anticipation of a public offering. Moreover, the operating agreement should discuss certain terms and conditions of a merger or conversion, such as the basis for converting membership interests into shares as well as their relative rights and limitations. This will provide the owners with leverage when working with future investors and should reduce future legal costs.

In most LLC statutes, unanimous member approval is required for mergers and conversions. Certain collective action problems, including the potential for deadlocks and hold-outs, are likely to exist in a ten member LLC. Therefore, the owners should incorporate a provision into the operating agreement reducing the approval required for a merger or conversion from unanimous consent to a supermajority. This is a more appropriate solution for preventing deadlocks and hold-outs than a buy-out provision because of the difficulties associated with identifying the appropriate buy-out valuation for an early-stage technology firm.

IV. AGENCY AND LIABILITY OF MEMBERS TO THIRD PARTIES

A. Limited Liability of the LLC

Technology firms may have high product or service liability risks from which the owners would like to insulate themselves. These risks could include general market risks, risks of intellectual property infringement suits, as well as risks associated with capital and debt financing. Therefore, the limited liability characteristic of the LLC is incredibly valuable. Generally, LLCs protect their members, managers and agents from personal liability for the debts of the LLC. [9] By contrast, partners generally are personally liable for debts of the partnership. For example, in Thompson v. Wayne Smith Construction Company, the court held that once the partnership assets were exhausted, the partner’s personal assets could be used to satisfy the debt owed to a creditor. [10] Partners sometimes try to contract out of personal liability in contracts with creditors, as was the case in both Regional Federal Savings Bank v. Margolis and Commons West Office Condos, Ltd. v. Resolution Trust Corporation, but these negotiations can be costly and the contracts are not always enforced by the courts. [11] By simply choosing the LLC form, the owners will insulate themselves from the risks in these cases.

B. Limiting Members’ Authority to Bind the LLC

In an LLC, an agency relationship exists between the members and the firm. [12] As an agent of the firm, the members have the authority to bind the firm in certain circumstances. The authority to bind stems from the agency principles of actual, apparent and inherent authority. [13] The owners of the LLC may want to modify provisions in the operating agreement to minimize the threat of liability in certain situations.

For example, standard provisions in most operating agreements define the purpose of the LLC broadly. While a broad statement of the LLC’s purpose increases the flexibility of the LLC, it also makes it easier for the members to bind the firm in undesired ways. [14] In Zimmerman v. Hogg & Allen Professional Association, for instance, the court stated that the scope of the partnership’s business was relevant to whether the firm was liable on a theory of apparent authority. [15] Similarly, in Heath v. Craighill, Rendleman, Ingle & Blythe, a client whose attorney made bad investments on behalf of the client could not recover his investment losses from the attorney’s law firm because the purpose of a law firm was not to process investments. [16] Therefore, the lawyer was therefore acting without apparent authority. [17] To reduce the threat of members’ binding the LLC in unforeseen ways based on apparent authority, the owners should narrow the purpose of the LLC in paragraph 2.4 of the CA to the creation and selling of software and related services.

Additionally, the management and control structure of the LLC can limit members’ ability to bind the LLC. In a member-managed LLC, each member has full authority to bind the LLC. In a manager-managed LLC, the non-managing members’ ability to bind the firm is reduced because their actual, apparent and inherent authorities are reduced. Therefore, by adding the management structure provisions previously described above defining roles like CEO, CFO, CMO, COO and CTO, the members may reduce the ability for the other members to bind the firm.

C. Limiting the LLC’s Liability for Members’ Torts

Under the theory of respondeat superior, as described in Millan v. Dean Witter Reynolds, Inc. and Mains v. II Morrow, Inc., a principal is not liable for the torts of an agent unless the conduct is within the agent’s “scope of employment.” [18] In Jackson v. Richter, for example, the court held in favor of an employer because the extramarital conduct of its employees was “outside the scope of employment” and not intended to serve the employers purpose. [19] While the members cannot draft around vicarious liability entirely, by narrowing the “purpose” of the LLC they can reduce their potential liability by making more activities fall “outside the scope” of the LLC’s objectives.

Furthermore, a principal is generally not liable for torts committed by independent contractors unless the activity is abnormally dangerous. [20] In Anderson v. Marathon Petroleum Co., for instance, the court found that sandblasting was not abnormally dangerous and that the employer was therefore not liable for the tort of its independent contractor. [21] Here, the owners of the technology LLC are likely to engage with independent contractors for numerous activities (e.g. telemarketing, press relations, website design, graphic design, etc.). Therefore, the owners should modify the operating agreement to make clear that anyone not constituting an “employee” as defined by the agreement should be considered an independent contractor.

Many LLC statutes permit indemnification of any member, manager or agent against any and all claims subject to the standards and restrictions set forth in the operating agreement. The owners may want to provide for broad indemnification of members for activities engaged in within the scope of their employment. Together with the provision narrowing the LLC’s purpose, the indemnification provision could serve to reduce potential liability for activities outside the scope of the LLC while encouraging risk-taking that is in the best interests of the firm. This is particularly important in small technology companies where calculated risks can make or break the firm’s success. Therefore, the operating agreement should be updated to state that the LLC will indemnify “any member or manager against any loss or threat of loss” resulting from legal claims “related to the performance of any act concerning the business or activities of the LLC” as long as the member acted “in good faith within what was reasonably believed to be the best interest of the LLC.” [22] Additionally, the indemnification provision should be written to cover the fees and costs of defending these actions.

D. Reducing the Risk of Liability Under Federal Securities Laws

Federal securities law, including the Securities Act of 1933 and the Securities Exchange Act of 1934, provide a wide range of anti-fraud and affirmative disclosure provisions that protect investors in connection with the purchase or sale of a “security.” Courts look to the form of the transaction, among other things, to determine whether or not the sale was of a security and therefore subject to federal securities law. In Rivanna Trawlers Unlimited v. Thompson Trawlers, Inc., the court held that general partnership interests were not securities and were not subject to the federal securities laws. [23] While some state statutes have held that LLC interests are securities, others look to whether the LLC is centrally-managed versus member-managed. Therefore, by granting members of the LLC substantial governance powers similar to a partnership or closely-held corporation, the technology firm will reduce the likelihood that its membership interests will be considered “securities” subject to federal securities laws.

V. PROPERTY RIGHTS OF MEMBERS

A. Prohibiting the Transfer of Membership Interests

Under the default rules of most LLC statutes, a member may assign his economic interest in the LLC but may not assign his management interest. However, a small technology firm may have incentives to restrict or potentially prohibit even the transfer of financial rights in the LLC. First, if a member could transfer his financial interest without losing his management interest, his incentive to shirk his management duties would be greatly enhanced. Second, even if the member lost his management interest, transferring a financial interest in the LLC to a third party that lacks management power would dilute the interests of the other owners of the firm. They would be left with a larger percentage of management responsibility with the same residual claim. That is, they would have to work harder for the same upside. Third, the member with only financial interests in the LLC will have different views than managing owners regarding extraordinary decisions such as mergers, acquisitions, or going to the capital markets.

Therefore, paragraph 9.1 of the CA should prohibit members from assigning both their financial and their management interests. A simple ban on transferring membership interests will prevent the issue in Sunshine Cellular v. Vanguard Cellular Systems, Inc in which ambiguity over the definition of the interests led to a court proceeding. [24] Moreover, a simple ban on interest transfers will reduce the likelihood of issues like in In re Asian Yard Partners in which the parties disagreed about the proper interpretation of a transfer restriction in the partnership agreement. [25] Additionally, paragraph 9.2 of the CA could be removed because assignees would not exist and therefore would have no rights. This rule would impact paragraph 10.2 on the consequences of dissociation. Language should be added to provide that a dissociating member loses all financial and membership rights which, upon member dissociation, would revert back to the remaining members in proportion to their existing equity ratios.

An alternative approach to a full ban on transfer would be to provide existing managers the right of first refusal to “buy back” the member’s interest. This would be similar to the right of first refusal granted to members of a limited partnership in Lawrence v. Cohn. [26] However, in Lawrence problems arose because unclear drafting made it ambiguous when the right of first refusal right was triggered. [27] Similarly, in Valinote v. Ballis, the court struggled with how to interpret a buy-sell clause in an LLC operating agreement when a dissociating member wanted a valuation of his interest. [28] Therefore, the valuation and interpretation problems associated with buy back provisions make this approach inferior to a simple ban on the transfer of LLC membership interests.

Another benefit of the simple ban on transferring interests is that it will help prevent the problem in Labovitz v. Dolan in which a general partner refusing to distribute cash flow to limited partners and squeezed them into accepting below book value for their interests. [29] In the technology LLC, the members will have no expectation of distributions prior to a major liquidation event such as an acquisition or IPO. Furthermore, the ban on interest transfers will make it difficult to squeeze members out. Members will know that if they leave the LLC early, they will lose their membership interests.

B. Intellectual Property: Tax Benefits of the LLC

The principal assets of a small technology firm may differ from other types of firms. A software vendor’s primary asset is likely to be its intellectual property. Technology firms frequently own patents on the technology and processes they have developed. Typically, the sale of patents will qualify for more favorable long-term capital gains if the patents are held by individuals rather than the firm. In corporations, patents must be held by the firm and therefore their sale generates ordinary income taxed more heavily. Individual members of the technology LLC are likely to be treated the same as partners and will likely qualify to hold patents individually and thereby benefit from more favorable long-term capital gains treatment. For clarity, language in the operating agreement should be drafted stating that all intellectual property is to be held individually by the respective members of the LLC.

VI. FIDUCIARY DUTIES AND REMEDIES

Fiduciary duties arise when an owner delegates power to manage his property to one who does not fully own the property and therefore lacks the owner’s incentive to maximize the property’s value. In corporations, for instance, the board of directors owes a fiduciary duty to the shareholding owners of the firm. Similarly, as described by the courts in Starr v. Fordham and Meinhard v. Salmon, partners and joint venturers owe a fiduciary duty to their co-partners and co-venturers. [30] In LLCs, members typically owe a fiduciary duty of loyalty and due care to the firm. [31] While fiduciary duties can help align the incentives of the agent with that of the principal, they may also discourage managers from taking risks for fear of having to pay for a mistake. In these situations, owners of firms may contract out of their fiduciary duties through partnership or operating agreements. For example, in Exxon Corporation v. Burglin, a general partner was not held to breach his duty because the partnership agreement unambiguously abrogated the duty of loyalty. [32] In the small technology LLC, the members’ ability to take risks may be critical to the firm’s success. Additionally, as both owners and managers, members are more naturally aligned with the interests of the LLC. Therefore, the owners of the technology LLC may want to reduce the fiduciary duties of due care and loyalty.

A. Limited Duty of Loyalty Except for Usurped Business Opportunities

By default, the owners will have a duty of loyalty to the LLC. [33] Generally, the operating agreement of the technology LLC should provide for a small duty of loyalty. However, the threat of usurped business opportunities may be of particular concern for the owners. The owners may want to prevent the other owners from operating opportunistically and usurping potential business opportunities for their own benefit rather than the benefit of the LLC. Breach of the duty of loyalty frequently involves usurped business opportunities. In Tri-Growth Centre City, Ltd. v. Silldorf, Burdman, Duignan & Eisenberg, the court held that it was a breach of fiduciary duty for a partner to use confidential information to bid against the partnership in buying property. [34] Similarly, in Triple Five of Minnesota, Inc. v. Simon the defendant breached his fiduciary duty of loyalty to the plaintiff by usurping a $700 million business opportunity. [35]

In the technology LLC here, the threat of usurped business opportunities may be greater given the narrow definition of the LLC’s scope in paragraph 2.4. Therefore, the owners of the technology LLC may want to include language in the operating agreement that reduces the threat of members usurping business opportunities. The operating agreement should be explicit in defining what business opportunities the members may or may not engage in. Additionally, the operating agreement should prohibit members from using confidential LLC information in ways that are not in the best interests of the LLC. Finally, the non-compete provision should also reduce the likelihood of members of the LLC usurping business opportunities.

B. Abrogation of Members’ Duty of Care

Most LLC statutes provide that the members have a duty of care to either refrain from reckless conduct or to act as a prudent person would in similar circumstances. [36] This means that partners are liable for certain mistakes in managing the firm even if they act unselfishly. In the technology LLC, where there is little separation between management and ownership, the duty of care is less important. With ten owners, there will likely be sufficient monitoring of individual members to prevent most breaches of care. Therefore, the members of the technology LLC should modify paragraph 7.1 of the CA to contract for no duty of care between the owners. The direct litigation costs of potential liability, as well as the indirect costs of constraining the members’ exercise of discretion probably outweigh the benefits in deterring bad decisions.

C. Arbitration as Sole Remedy for Breaches of Fiduciary Duties

Per Sertich v. Moorman, litigation over breaches of fiduciary duty can be costly and time-consuming. [37] Therefore, the operating agreement should waive any default remedies for breach of fiduciary duty and include a broad arbitration clause providing that these claims are to be arbitrated out of court. Additionally, members may want to get rid of an errant or incompetent member from time to time. Therefore, the operating agreement should provide that a finding that fiduciary duties have been breached is grounds for expelling the breaching member, pursuant to a supermajority vote of the other members. An expulsion provision will prevent the issue in Cadwalader, Wickersham & Taft v. Beasley, in which the court held that a partner could not be expelled where the partnership agreement did not provide conditions for expulsion. To mitigate the risk of opportunism, the operating agreement should specify that expulsion of a member for breach of fiduciary duty may not be done in bad faith.

VII. DISSOCIATION AND DISSOLUTION

A. Dissolution and Liquidation

If the technology LLC is unable to get to market with a compelling set of software products, it may be forced to dissolve. Issues frequently arise in determining how to distribute the firm’s assets upon dissolution. For example, in Kessler v. Antinora, the court struggled with identifying the appropriate way to allocate the debts of a general partnership where one partner contributed money and the other contributed labor. [38] To avoid ambiguity, the operating agreement should specify how the assets of the firm will be distributed upon dissolution. While liquidation rather than in kind distribution of assets is common, [39] it may not be appropriate in a small technology firm. It is likely that upon dissolution the technology LLC will have little cash. Its primary assets will be its intellectual property and its physical assets. The liquidation value of the firm’s intellectual property may be incredibly low. The owners may prefer other distribution options for the intellectual property, such as providing that each owner receives full rights to it, auctioning it off to one of the ten owners, or requiring that it be made available to the public as open source code. Therefore, language should be added to paragraphs 11.1 and 11.3 of the CA allowing members to make distributions upon dissolution based on a flexible dissolution agreement ratified by a supermajority of the members at or before dissolution. In lieu of an accepted dissolution agreement, the assets should be liquidated. This approach should also be used for the physical assets of the LLC, many of which will be pieces of technology (i.e. servers, laptops) whose liquidation value is likely to be low relative to their in-kind value to the members. Therefore, paragraph 11.1 of the CA should be modified to specifically authorize in kind distributions.

B. Effect of Member Dissociation

There are many ways that a member of an LLC may dissociate. The member may voluntarily dissociate or may be expelled based on events agreed to in the operating agreement, such as breach of fiduciary duties, excessive shirking, or maintaining a negative capital account. Moreover, dissociation may occur as a result of the death of a member. A major benefit of the LLC form is that the LLC is a considered a distinct legal entity and therefore member dissociation need not force dissolution.

Each member of the technology LLC has a strong incentive to encourage the other members not to leave the LLC. Innovative technology is costly to create but easy to copy. A dissociating member will leave with information about the practices and procedures that the LLC has developed that could be threatening to the LLC. Therefore, paragraph 10.3 of the CA defining the distribution rights of dissociating members should be redrafted to create additional incentives for the members not to dissociate. For example, language should be added stating that dissociating members have no right to a return of their capital contributions and no right to positive balances in their capital accounts. [40] In Darr v. D.R.S. Investments an issue arose when partners wanted a dissociating member to pay his negative capital account balance even though the partnership assets as a whole had appreciated in value. [41] The technology LLC could avoid the problem in Darr by stating that dissociating members are not liable for a negative balance. To discourage members from being delinquent in their accounts, maintaining a negative balance could constitute grounds for expulsion. This would be similar to Chandler Medical Building Partners v. Chandler Dental Group in which the partnership agreement provided that a partner could be treated as withdrawn in the event of failing to maintain his capital account in accordance with his pro-rata responsibilities. Providing in the operating agreement that a dissociating member has no right or obligation to any excess or deficiencies in their capital accounts, or to a return of their initial contribution, should reduce ambiguity and encourage members to stay with the firm.

To further discourage voluntary dissociation, paragraph 10.3 should provide that dissociating members have no right to any of the goodwill or going concern of the LLC. [42] Dissociating members will lose their membership interests and these interests will be distributed among the remaining members proportionally based on their current equity ratios. This approach solves the problem of valuating goodwill that occurred in Spayd v. Turner in which the court held that a dissociating lawyer was entitled to an accounting that included a portion of the firm’s goodwill. [43] Valuation of the going concern of the small technology LLC will be incredibly difficult, particularly in the early years when the firm is likely to have low revenues and high research and development costs. The “fair market value” of a dissociating member’s interest would be difficult and costly to estimate. [44] Therefore, by preventing a dissociating member from having a right to a portion of the LLC’s going concern, the members can both discourage dissociation and avoid potentially costly litigation or arbitration proceedings to determine a valuation of the interest. Similarly, to protect the LLC in the event of a member’s death, the owners should modify paragraph 10.1 of the CA to make clear that the financial and management interests of the deceased revert immediately back to the LLC.

Moreover, the primary asset of the LLC is likely to be its intellectual property. Therefore, paragraph 10.2 of the CA should add language making clear that a dissociating member has no rights in the intellectual property of the firm. The language should specifically prohibit the dissociating member’s use of the firm’s intellectual property and trade secrets. Moreover, members of the LLC may fear that a dissociating member will attempt to compete with the LLC. A dissociating member’s skillset and knowledge gained from working at the LLC will create strong incentives to compete. Competition is often an issue on dissociation. In Meehan v. Shaughnessy, for instance, a dispute arose over whether law partners could secretly solicit clients from a firm while they were still employed. [45] Additionally, the court in Howard v. Babcock held that an agreement among law partners that imposed a toll on departing partners who competed with the firm was enforceable. [46] Therefore, the owners of the technology LLC should include language that reasonably restricts the dissociating member’s ability to compete with the LLC.

 

Posted in Law and Technology | No Comments »

A Critique of Brand X v. FCC

January 12th, 2007

CASENOTE: A CRITIQUE OF BRAND X v. FCC

I. Introduction

The Telecommunications Act sought to eliminate obstaclesthat new telecommunications companies had in order to stimulate and promotecompetition in these markets. The statute maintained significant obligations toproviders of telecommunication services but left providers of informationservices subject to much less stringent regulation. Initially, the FCC did not takea position on the regulatory classification of cable modem services. However, afterfederal courts began classifying cable modem services in different ways, theFCC issued a declaratory ruling intending to develop a national legal standard.The new ruling classified cable modem services as purely “informationservices.” Despite the FCC ruling, in Brand X the circuit court upheld conflictingcircuit precedent stating that cable modem service was a combination ofinformation services and telecommunication services. Brand X Internet Servs.v. FCC, 345 F.3d 1120 (9th Cir. 2003) at 1132. If the circuit court hadfollowed the FCC’s classification, cable operators would have been able to leveragethe less stringent provisions associated with information services, sparkingcompetition. Instead, the court adhered to its prior precedent, inhibiting newcompanies from entering the market and stopping this policy debate in itstracks. Id. at 1133.

The purpose of this casenote is to analyze the impact andadequacy of the holding in Brand X. Part II of this casenote traces thedevelopment of law and provides the background necessary to explain the impactof Brand X. Part III identifies the key issues in Brand X andthe rationale behind the court’s opinions. Part IV analyzes the outcome of thecase including whether the court applied the appropriate rule of law andwhether the issues presented were adequately resolved.

II. Background

To understand the impact of the court’s holding in BrandX, it is essential to first understand how the relevant case law hasevolved.

A.Chevron:Two-Step Rule

Prior to Chevron, a court was free to impose its ownconstruction on an ambiguous, agency-administered statute. Richard L. Pierce,Jr., Reconciling Chevron and Stare Decisis, 85 Georgetown L.J. 2225, 2225(1997). However, in Chevron the Supreme Court greatly restricted theability of courts to impose their own constructions of these statutes. Chevron U.S.A., Inc. v. NRDC Inc., 467 U.S. 837 (1984). The EPA was tasked withadministering the Clean Air Act and it stated that States could treat all ofthe pollution-emitting devices within the same industrial grouping as thoughthey were in a single ‘bubble.’ Id. In deciding what deference to giveto EPA’s rule, the Court defined a two-step test. Id. Step one said thatif Congress had spoken to the precise question at issue then there was noambiguity and the court had to follow Congressional intent. Id. at 842.Step two said that if the statute was silent or ambiguous with respect to theissue, the court had to uphold the agency’s interpretation if it was a “permissibleconstruction” of the statute and was not “arbitrary, capricious or manifestlycontrary to the statute.” Id. at 843.

The Chevron test required a court to uphold anagency’s interpretation even if the court would have interpreted the ambiguitydifferently. Id. at 844. The rationale for giving this level ofdeference to agency interpretation was that an agency is better positioned tounderstand the complexities associated with the statutes they are entrusted toadminister. Id. at 845.

In Chemical Waste, the court applied the Chevrontest. Chemical Waste Management, Inc. v. U.S. Environmental ProtectionAgency, 873 F.2d 1477 (D.C. Cir. 1989). Under the Resource Conservation andRecovery Act (RCRA), operators of hazardous waste treatments being assessedpenalties for non-compliance had a right to a “public hearing.” Id. at 1478. The EPA’s interpretation of “public hearing” did not require that formaladjudication proceedings be given in all circumstances. Applying step one ofthe Chevron test, the court held that Congress had not spoken directly tothe meaning of “public hearing” in the RCRA. Id. at 1480. Applying step twoof the test, the court held that “public hearing” was ambiguous and that theEPA’s interpretation was a permissible construction of the statute. Id. at 1482. The court therefore upheld the EPA’s interpretation of the statute. Id.

In Satellite Broadcasting, the Copyright Office’sinterpretation of the term cable system conflicted with circuit precedent. SatelliteBroadcasting & Communications Ass’n of Am. v. Oman, 17 F.3d 344 (11thCir. 1994). The circuit precedent held that satellite carriers were cablesystems and were therefore entitled to subscribe to the compulsory licensingscheme entitling them to copyrighted television broadcasts in exchange forpaying royalties. Id. at 346. Afterwards, the Copyright Office modifiedits regulations finding that satellite carriers were not cable systems. Id. at 347. Applying the Chevron test, the court overruled prior precedentstating that they would follow the Copyright Office’s construction because itwas not arbitrary, capricious or contrary to the statute’s clear meaning. Id. at 347.

B.Neal:Defending >StareDecisis

After Chevron, it was unclear whether Supreme Courtprecedent always trumped new agency interpretations. In Neal, theSupreme Court addressed this issue stating that a court must follow SupremeCourt precedent, regardless of conflicting agency regulations. Neal v. United States, 516 U.S. 284 (1996). In Neal, the Court had to decide how to calculate theweight of LSD for the purposes of calculating the mandatory minimum sentencedefined in the Anti-Drug Abuse Act. Id. Based on prior Supreme Courtprecedent, the calculation should include the weight of the blotter paper,regardless of the type its material. Id. 289. However, the United StatesSentencing Commission subsequently published a new way to calculate the weightof LSD. Id. 287. The Supreme Court held that its prior precedent and thedoctrine of stare decisis trumped the agency’s later interpretation ofthe statute. Id. 296. The Court rationalized this rule stating thatCongress, not the Supreme Court, has the responsibility for revising itsstatutes. Id.

While Neal made clear that Supreme Court precedenttrumped conflicting agency interpretation, courts disagreed about whether thismeant that circuit precedence did. Aguirre and Bankers Trustdemonstrated this confusion. Aguirre v. INS, 79 F.3d 315 (2d Cir. 1996);Bankers Trust N.Y. Corp. v. United States, 225 F.3d 1368 (Fed. Cir.2000). In Aguirre, the circuit court overruled circuit precedent givingdeference to the Board of Immigration Appeals rule that an alien could still beconsidered for waiver of deportation if his felonious state law offense was nota felony under federal law. Aguirre, 79 F.3d 315. The court interpretedthe rule in Neal to be that an agency cannot compel a court to forgo theprinciples of stare decisis, but that a court can voluntarily acceptsuch guidance. Id. at 317.

In Bankers Trust, on the other hand, the FederalCircuit Court of Appeals held the Court of Federal Claims erred in givingdeference to a new regulation from the IRS over existing precedent. BankersTrust N.Y. Corp., 225 F.3d 1368. The circuit court argued that thedoctrine of stare decisis should probably be even stronger at the circuitcourt level, since because poor judicial interpretations of statutes would becorrectable not only by Congress, but also by the court itself sitting en banc,as well as by the Supreme Court. Id. at 1375. The court therefore upheldits prior precedent despite the IRS’s conflicting interpretation. Id.

C.Mead: NoDeference without Delegation

In Mead, the Supreme Court further limited theapplication of the Chevron test. United States v. Mead Corp.,533 U.S. 218 (2001). It held that an agency’s interpretation of a statutoryprovision only qualifies for Chevron deference when Congress explicitlyor implicitly delegates to the agency the authority to make rules that wouldcarry the full force of law. Id. In his dissent, Scalia arguedthat this holding modified the presumption in Chevron that agencies can resolveambiguity in the statutes they have been authorized to enforce to a presumptionthat they can not. Id. at 239. He argued that this would inhibit theevolution of statutory law because once a court had spoken it would be “unlawful”for the agency to take a contradictory position. Id. at 247. He notedthat this would be particularly “bizarre” when it occurs because an agencyfailed to address the issue before it was presented to the court. Id. at 247. Short of Congressional legislation, there would be no way for the agencyto modify the court’s rule. Id.

However, Heimmerman showed that the rule in Meaddid not inhibit the evolution of statutory law in situations where Congressspecifically delegated its rulemaking power to the agency. Heimmermann v.First Union Mortg. Corp., 305 F.3d 1257 (11th Cir. 2002) In Heimmerman,the circuit court gave deference to interpretations of the Real EstateSettlement Procedure Act (RESPA) made by the Department of Housing and UrbanDevelopment (HUD). Id. The court held that RESPA expressly authorizedHUD to prescribe rules and regulations necessary to further the purposes of thestatute. Id. at 1261. Therefore, the rule from Mead was satisfiedand the application of the Chevron test was appropriate. Id.

While Mead limited the application of the Chevrontest only to situations in which the agency was specifically delegatedrulemaking power by Congress, it did not address the question left open in Nealof whether circuit precedence trumped subsequent agency interpretation.

III. Discussion

In Brand X, the court waspresented with the issue of whether circuit precedent trumped subsequent agencyinterpretation. Brand X Internet Servs. v. FCC, 345 F.3d 1120 (9th Cir.2003). The circuit precedent stated that a cable modem service was acombination of information services and telecommunication services whereas asubsequent FCC ruling stated that it was solely an information service. Id. The significance of the distinction was that the Telecommunications Actsubjected providers of telecommunication services to significant obligationswhile subjecting providers of information services to much less stringentregulation. Id. at 1126. Each of the three justices on the panelprovided a separate opinion explaining their rationale for upholding circuitprecedent. Id. This section of the casenote explores these threeopinions in more detail.

A.MajorityOpinion

In Brand X, the majorityopinion upheld the classification defined in its circuit precedent despite theconflicting FCC ruling. Id. at 1132. While the majority opinion did notreference the Mead rule directly, it did acknowledge that the FCC was chargedby Congress with the administration of the Telecommunications Act and that Congresshad delegated its authority to enforce the provisions of the Act. Id. at 1127. The court worked from the premise that federal courts are bound bytheir prior holdings under the doctrine of stare decisis. Id. at 1130. The court said that an exception to stare decisis existed in itscircuit where precedent conflicted with subsequent agency interpretation. Id. However, it stated that the exception did not apply where the precedent representedthe only possible interpretation. Id. at 1131. The court then arguedthat the exception did not apply to Brand X without arguing that itsprevious interpretation of cable modem services was the only possibleinterpretation. Nevertheless, after dismissing this exception, the court then lookedat the rule in Neal which stated that once the Supreme Court determineda statute’s meaning, it must adhere to its ruling under stare decisisregardless of an agency’s subsequent interpretation. Id. The majorityopinion then held that they were bound by their prior precedent because therewas nothing in Neal to suggest that this rule only applied to theSupreme Court. Id. at 1132.

B.O’ScannlainConcurring

In his concurring opinion, JusticeO’Scannlain took a different approach to analyzing the case. O’Scannlainargued that the majority opinion, though correct, was strikingly inconsistentwith Chevron’s underlying principles. Id. at 1132. O’Scannlainargued that the precedent case “in essence, beat the FCC to the punch” becauseit was forced to rule on the meaning of “cable modem service” before the FCCitself could rule on it. Id. at 1133. He said that this case representedthe “bizarre” circumstance Scalia envisioned in his dissent in Meadwhere the evolution of statutory law would be inhibited by judicial precedent. Id. at 1134. Nevertheless, he concurred with the majority because he believed thecourt’s precedent was controlling. Id.

C.ThomasConcurring

In his concurring opinion, JusticeThomas analyzed the case using the Chevron test without first addressingthe limitations placed on Chevron by Neal and Mead. Id.Thomas spent the bulk of his analysis explaining why its circuit precedentconcluded that a cable modem service is both an information service and atelecommunication service. Id. He found that the definitions used in theTelecommunications Act and the overall legislative context suggested that Congressclearly intended a cable modem service to be part telecommunication service. Id. at 1140. He argued that even without any precedent, the court would have to cometo the same holding because the terms used by Congress had clear meaning. Id. at 1140. Therefore, it appears Thomas found that step one of the Chevron testwas not satisfied because the agency interpretation was in conflict with theclear intent of Congress. Id.

IV. Analysis

This section analyzes whether thejustices in Brand X used the appropriate test, whether they applied thetest correctly in reaching its conclusion, and whether the resulting holdingwas adequate given its practical implications and relevant policyconsiderations.

A.The Proper Test

The circuit court in Brand Xhad to determine whether the principles of stare decisis compelled it tofollow circuit precedent or whether it should have given Chevrondeference to the conflicting ruling from the FCC. Given the legislativehistory described in Part II, the proper test for the court to follow would havebeen to first apply the test in Mead. If the court found that Congresshad not expressly authorized the FCC to prescribe rules and regulationsnecessary to further the purposes of the Telecommunications Act, the courtshould have followed circuit precedent. If the court found the FCC had beengranted this authority, it should have applied the rule in Neal. While Nealwas clear that Supreme Court precedent trumps subsequent agencyinterpretation, it left open the question of whether circuit precedent does. The court in Brand X should have addressed this issue. If it found thatNeal was limited to Supreme Court decisions, the circuit court shouldhave applied the two-step Chevron test. Otherwise, it should haveupheld circuit precedent.

B.Was the Test Applied Properly?

The majority opinion did notproperly apply the above test. While it did acknowledge that the FCC has theappropriate authority to regulate the Telecommunications Act, it did notreference this as being a requirement of Mead. Next, rather thanapplying Neal to determine whether the rule of stare decisisshould apply when circuit court precedent conflicts with subsequent agencyinterpretation, the court first assumed stare decisis applied andanalyzed whether there were any exceptions within the circuit. After thisdigression, the court correctly recognized the issue left open in Nealbut failed to adequately address it. The court’s only logic for holding that Nealshould be extended to cover circuit precedent was that there was nothing in Nealthat suggested it shouldn’t be extended. The opinion needed to justify itsstance on this issue by talking about the pros and cons of interpreting therule in Neal in this way. Additionally, the court should have explainedthat, given its interpretation of Neal, the Chevron test was notrelevant because they were compelled to follow circuit precedent. The courtdid not address Chevron at all.

O’Scannlain’s concurring opinionalso misapplied the proper test. While he recognized that the first step wasto start with Mead he applied the rule in Mead incorrectly. O’Scannlain believed this case was an example of the bizarre circumstancesScalia referred to in his dissent in Mead. However, these circumstancesonly occur when Congress has not explicitly or impliedly delegated rulemakingauthority to the agency. In Brand X, it was clear that Congress haddelegated this authority to the FCC. Therefore, Scalia’s concerns were notrelevant. O’Scannlain did not even address the issues of stare decisisin Neal or the two-step test in Chevron after misapplying therule in Mead.

Thomas also misapplied the propertest in his concurring opinion. First, he did not discuss Mead at all. However, he may have simply recognized that Congress had explicitly granted theFCC authority to regulate the Telecommunications Act and that Mead wastherefore satisfied. Next, he failed to discuss the issue of whether Nealapplies to circuit court precedent. It is possible that Thomas simply assumedthat Neal was limited to Supreme Court precedent. However, given thatsome other circuits had held that Neal did apply to circuit precedent,and the majority opinion in Brand X even claimed that it did, Thomasshould have explained his reasoning for limiting Neal to Supreme Courtprecedent. Next, Thomas applied the Chevron test finding that step onewas not satisfied because the statutory definitions combined with thelegislative context compelled a single interpretation of the meaning of cablemodem service. Therefore, he correctly found that Congress had spoken to theissue and that the court had to follow Congressional intent, regardless of agencyinterpretation.

Therefore, none of the justicesapplied the proper test in arriving at their conclusions.

C.Was the Outcome Adequate Given Policy Considerations?

The outcome of Brand X isnot adequate in light of its practical implications and relevant policyconsiderations. The primary purpose of the Telecommunications Act was toeliminate obstacles that new telecommunications companies had in order tostimulate competition in telecommunication, information and cable services. The FCC is much better positioned than a circuit court to determine what definitionof cable modem service is best to stimulate competition in the areas it hasbeen charged with administering. By upholding circuit precedent over the FCC’ssubsequent ruling, the circuit court in Brand X inhibited the naturalevolution of telecommunications law. While the court suggests this isCongress’s responsibility, the practical reality is that agencies have agrowing responsibility for interpreting Congressional legislation. This rulingprevents the FCC from proactively adapting their interpretation of theTelecommunications Act in ways that help further the goal of stimulatingcompetition. Rather, this ruling puts unnecessary burden on Congress to modifythe act as necessary over time which will be a much slower process. Given thatneither courts nor Congress could have the ability to anticipate all potentialproblems with the Telecommunications Act, this ruling will force Congress toconstantly attempt to fix problems that could have been anticipated by agencieswith more time and resources dedicated to this purpose. This is a particularlylarge problem in areas like telecommunications and information services inwhich advances in technology occur incredibly quickly and drastically alter thecompetitive landscape.

Moreover, given that differentcircuits could arrive at different conclusions as to the meaning of cable modemservices, the ruling in Brand X is likely to prevent the consistentapplication of the law. Cable modem service providers will be subject todifferent standards and rules depending on where they are brought to courtwhich will encourage horizontal forum shopping. Similarly, this will inhibitthe FCC’s ability to provide a coherent telecommunications policy because differentsections of the Telecommunications Act will be interpreted by the judges withdiffering views creating binding precedent that the FCC will be unable toadjust. While this ruling will put more pressure on Congress to create clearerlegislation, the reality is that this will simply pile more legislativecomplexity onto an Act that is already confusing and self-contradicting.

Even worse, this decision does notprotect the reliance interests of new companies providing these types of cablemodem services. This type of company will spread across geographical barriers providingservices as part of a large, interconnected network. However, this ruling willprevent these companies from relying on FCC rulings because they will neverknow when some jurisdiction has contradicting precedent. This will exposethese small companies to different rules in different areas creating a cost-prohibitive,administrative nightmare.

Nor does this ruling protect theinstitutional reputation of the American judicial system. In Brand X,Justice O’Scannlain himself said that this rule was “bizarre.” The moredecisions from American courts that the public considers bizarre and contraryto reason, the less respect they will have for the judicial institutionsthemselves.

V. Conclusion

The holding in Brand X wasbased on a misapplication of the law on the part of all three Justices. Moreover, the holding is contrary to the original purpose of theTelecommunications Act of stimulating competition and reducing barriers to entryin these markets.

Posted in Law and Technology | No Comments »

Google Book Search, BitTorrent and Copyright Law

January 12th, 2007

GOOGLE BOOKSEARCH, BITTORRENT AND COPYRIGHT LAW

INTRODUCTION

Emerging internet-basedtechnologies have created new issues in copyright law. Courts increasinglyhave to address complex questions involving the balance of encouraginginnovation while protecting the unique expressions of artistic creators. Twotimely examples involve file-sharing applications and internet search engines. File-sharing is the practice of making files available for users to downloadover the Internet.[1] A search engine is an application that helps users find and retrieveinformation stored on the Internet. [2] While file-sharing is a legal technology with many legal uses, most users use thetechnology to download copyrighted materials including music and movie files. Despite the users being the ones committing the infringements, under certaincircumstances companies creating the file-sharing applications themselves havebeen held liable under theories of contributory or secondary infringement. Regarding search engines, these companies often store cached versions ofcopyrighted web content in their search databases and index files and presentthis content to users in their search results. The movie, music, and bookindustries have engaged in significant litigation to try to curtailinfringements of their copyrights by file-sharing and search enginetechnologies.

This comment will discuss howexisting copyright law would likely be applied to two high-profile newtechnologies, BitTorrent and Google Book Search. BitTorrent is a new P2Pnetwork technology that should not be liable for the copyright infringements ofits users based on a theory of contributory infringement for distributing itsproduct or on a theory that it induced its users to commit infringement. GoogleBook Search, on the other hand, probably does not constitute a fair use ofcopyrighted material and therefore should be liable.

I. TECHNICAL BACKGROUND

A. P2P File-sharing Technologies

1. Introduction to P2P File-SharingTechnologies

File-sharing is the practice ofmaking digital files available for users to download over the Internet. [3] File sharing often uses a peer-to-peer model (P2P) in which files are stored onusers’ personal computers. [4] The P2P model provides a number of advantages for sharing files includingeliminating the need for server storage space and enabling faster filedownloads because the file may be available on many users’ computers. [5]The file sharing network consists of the protocols for how the sharing of fileswill take place. [6] Typically, users connect to the network using software programs installed ontheir own machines. The combination of 1) popular files, 2) ease of copyingdigital media, 3) high-bandwidth Internet access have made file-sharing overP2P networks explode. [7]

Napster was one of the first majorP2P software applications and was available for download for free. [8] Restricting file-sharing to only MP3 music files, Napster relied on centralizedservers for storing an index of all the MP3 files shared by users logged intothe network. [9] After registering,a user could access and search the index and then download the files directlyfrom one of the other users’ computers. [10] Napster was successfully shut down through legal attacks from the recordindustry and has since reemerged as a commercial online service. [11]

After the collapse of Napster,other P2P networks and P2P client software applications emerged includingAimster. [12] Aimster leveragedAmerica Online’s instant messaging application (AIM) to allow users to seetheir friend’s file listings and share files using the AIM network. [13]It consisted of a proprietary software client that could be downloaded forfree. [14] Users of Aimstercould leverage AOL’s network of users and search for file’s shared over thatnetwork. [15] Aimster’s serverwould search the computers of Aimster users to find the file requested andwould instruct the computer to transmit the file to the requester. [16]

Grokster and StreamCast wereadditional P2P client applications that became popular after the collapse ofNapster. They attempted to avoid liability for direct copyright infringement, inpart, by decentralizing their architectures. Grokster’s client applicationlicensed the FastTrack network technology. [17] StreamCast developed a similar product but relied on the Gnutella network. [18] Therefore, Grokster and StreamCast were new client software applications usedto connect to FastTrack and Gnutella’s decentralized P2P networks to enabletheir users to share files.[19] These networks differed from Napster in two primary ways: 1) users could searchfor any file type (they were not restricted to MP3s like in Napster), and 2)the network was decentralized (no centralized index servers). [20] The decentralized network architecture prevented any single broken link, suchas Napster’s index servers, from bringing down the network.[21]Grokster removed the need for centralized index servers through its use of supernodes. [22] Whenever a powerful computer with a high speed connection running Groksterconnected to the FastTrack network, it automatically became a supernode andacted as a temporary indexing server for other clients on the network. [23] It was believed that the decentralized nature of the network would insulate itfrom threats of direct copyright infringement.

Other P2P networks have alsoemerged. The eDonkey network emerged supporting the ability for users to beginsharing segments of files before they had been fully downloaded. [24] This speeds up the distribution and sharing of files on the network.[25] Moreover, eDonkey applications used to connect to the network often employedrating systems in which users who shared more files were rewarded with prioritywhen downloading. [26] In 2004, the eDonkey network became the most widely used file sharing networkon the Internet. [27]

2. BitTorrent – A Next Generation P2PFile-Sharing Application

BitTorrent is the name of a P2Ptechnology created by Bram Cohen in 2002.[28] BitTorrent has proven to be very adept at sharing large files such as movies,television shows and software applications. [29] Cohen says that he developed BitTorrent because his friends in the “jam band”community needed a better way to legally download entire concerts rapidly. [30] Traditional P2P technologies were too slow and unreliable.

The name “BitTorrent” refers to thenetwork protocol, the original application used to connect to the network, andthe .torrent file type used for sharing files. [31] To share a file using BitTorrent, a file is broken up into smaller fragments. [32] A .torrent file is created that provides information on how to reconstructthese fragments as well as information on how to find the relevant filefragments on the network. [33] Once a fragment of a file has been downloaded, that fragment becomes immediatelyavailable to be downloaded by other clients. [34] This architecture helped BitTorrent address one of the common problems withconventional file sharing applications. Namely, with conventional P2Ptechnology, high demand for a particular file could lead to bottlenecks,particularly for very large files that took a long time to download. [35] It took longer for larger files to spread through the network because each userhad to download the entire file before they could share it with others. [36] BitTorrent’s approach of breaking files into process fragments meant that userscould share portions of the large files faster, relieving the stress on theoriginal host server. [37] Similarly, while nodes in other file sharing networks share and download largenumbers of files making the bandwidth for each file smaller, users ofBitTorrent usually concentrate on transferring a single file or a smallcollection of files that make these transfers faster. [38]

One primary difference between theBitTorrent network and most other P2P networks is that the original clientapplications could not search the web for .torrent files. Therefore, to shareand download files, a user had to find the .torrent files themselves.[39] As a result, websites like Suprnova.org emerged that distributed .torrentfiles. While sites like Suprnova did not host the copyrighted filesthemselves, they provided access to the .torrent files for copyrighted music,movies, software applications and computer games. [40] Sites like Suprnova essentially became cataloguing tools for users to find anddownload copyrighted digital content and were quickly shut down. [41]The primary legal question is whether BitTorrent itself is liable for thedirect infringements of its users based on a theory of contributory orsecondary liability.

B. Introduction to Search Engine Technologies

1. The Evolution of SearchEngine Technologies

A search engine is an applicationthat helps users find and retrieve information stored on the Internet. [42] Search engines take in search criteria from users, typically in the form ofwords and phrases, and returns search results that match those criteria. [43] A search engine works by storing content on a large number of web pages anddelivering search results to users based on search criteria. [44] Search engines achieve this goal by 1) Crawling, 2) Indexing, and 3) Searching. [45] Search engines send out crawlers, or web spiders, which retrieve pages from theweb and follow every link that it finds. [46] The content on each page discovered is then analyzed and indexed based on itscontent. Search engines employ these indexes in order to ensure fast responsetimes. [47] This content isstored in the search engine’s index database. [48] Moreover, some search engines “cache” the pages they crawl storing all of the page’scontent, or some portion of the content, in the database as well. [49] These cached pages can be made accessible directly from the search engine’swebsite without hitting the original website. [50]

After a user has entered a query, thesearch engine looks for these “search criteria” in its index using a variety ofdifferent styles of algorithms. It then displays the resulting web pages andcontent that best matches the input criteria. Often these search resultsinclude other information about the returned content such as its location, adescription, some subset of the content, etc. The quality of a search enginedepends on its speed at returning results and the relevancy of those results tothe user. [51] Part of this is afunction of the matching algorithm and part of it is based on the amount ofinformation that is indexed. [52]

In 2001, the Google search enginebecame popular because its PageRank algorithm was very adept at quickly providingsearch results that were relevant to the user. [53]

2. Google Book Search

Google currently plans to scan thefull text of books into its search database as part of its Google Book Searchinitiative. [54] The initiative isbroken down into two key areas: the Print Publisher Program and the PrintLibrary Program. [55] Under the Print Publisher Program, Google obtains authorization from copyrightholders to scan the text of copyrighted books into Google’s search engine. [56] When a user enters a query into the Google search engine, the engine scans itsindexes and returns the user a set of bibliographic information on books thatmatch the user’s search terms. [57] Included with the search results are a few pages of the book surrounding thetext matching the user’s search criteria. [58] Links are made available so that the user can purchase the book from thepublisher or other online retailers. [59] The Print Publisher Program does not appear to raise any copyright issuesbecause Google obtains a specific grant to the content directly from thecopyright holder. [60]

Under the Print Library Project,however, Google has partnered with libraries from major educationalinstitutions like Harvard, Oxford and Standford to scan the text of the bookswithin those libraries without the permission of the original publishingcompanies. [61] In exchange forproviding the libraries with a digital copy of the materials, Google thenindexes a second digital copy for use within its search engines. [62] When a user enters search criteria that match information contained in one ofthese books they provide results that include a few sentences of surroundingtext. [63] Therefore, forbooks that are still subject to copyright restrictions, Google restricts theinformation provided to users. [64] Under this program, Google has also instituted an “opt out” policy in which thepublisher can provide Google a list of books that it does not want them to scanfrom the libraries’ archives. [65] Therefore, publishers are given the opportunity to prevent any of theircopyrighted materials from being made available through Google’s searchtechnology. [66] The primary legalquestion is whether Google’s Print Library Project violates copyright orwhether it is protected as a fair use of the copyrighted work.

III. LEGAL BACKGROUND

A. Copyright Law

1. Introduction to Copyright Law

While the term “copyright” does notappear in the United States Constitution, it is a right that Congress isempowered to secure. [67] Article I, Section 8 of the Constitution gives the United States Congress the power “[t]o promote the Progress ofScience and useful Arts, by securing for limited Times to Authors and Inventorsthe exclusive Right to their respective Writings and Discoveries.” [68] While Congress first exercised this power when it passed the Copyright Act of 1790, copyright law has been updatedsignificantly since then. [69] For example, the basis of modern copyright law in the United States is the Copyright Act of 1976. [70]

Essentially, copyrights are anexclusive set of rights granted by the government to creators of creative orartistic work and subject matter. [71] These rights enable the creators of creative work to regulate how these worksare expressed. [72] Among others, these rights include the right to 1) produce copies of the work,2) display the work publicly, and 3) sell copies of the work. [73] By maintaining these exclusive rights, the copyright holder can prevent othersfrom exercising these rights without their consent. [74] In this way, copyright has been considered a “negative right” because it can beused to prevent people from doing something. [75] Therefore, copyright is a means of protecting intellectual property whetherthat property takes the form of literature, movies, music, web content ortelevision. [76] Copyright law isprimarily geared towards protecting the form of the expression of the materialrather than the underlying idea itself. [77]

2. Copyright in an InformationEconomy

Many people believe that copyright,as it exists today, must change drastically in order to support contemporaryadvances in technology. People argue that copyright grants rights for too long(often the author’s life plus fifty years). [78] Moreover, differences in international copyright law are creating problems inthe new information economy in which the free flow of digital informationeasily transcends geographic and sovereignty borders. [79] Restricting the flow of this content into the “public domain” prevents would becreators from building on existing content and published works. [80]

The primary issue is thatcopyrighted works (books, movies, television shows, music, etc.) are easilycopied into digital form and available through file sharing and searchtechnologies over the internet. [81] Copyrights are often transferred from the original creators to record companiesand media conglomerates whose marketing and production resources increase thepotential audience for the works. [82] In exchange, the creators of creative work receive royalties or othercompensation. These large record companies and media conglomerates have becomemassive organizational structures whose assets often include the collection ofa large number of copyrights. [83]The worry of these large media conglomerates is that the Internet hasfacilitated the widespread sharing of files and content creating significantlosses in their industries. Others argue that copyright as a means of feedinglarge corporate entities is unnecessary and unjustified given that the creatorsof copyrighted material continue to produce new works. [84] According to Pew/Internet research, artists are “not deeply concerned” aboutthe file-sharing that happens online. [85] They want control over their creations, but most do not say internet piracy isa big threat. [86] Only 28% ofartists consider file-sharing to be a major threat. [87] Two thirds said that file-sharing poses a minor threat or no threat at all. [88] 83% have provided free samples of their work online and significant numbers saythat free downloading has helped them sell CDs and increase the crowds atconcerts. [89]

Therefore, in the digital agecopyright holders are finding it harder and harder to maintain theirmonopolistic control of content as millions of internet users find new,creative ways to share and distribute content. Given the difficulty of suingeach individual Internet user who uses file sharing and search technology tobreach copyright laws, copyright holders have started suing the distributors ofsuch technology under theories of secondary liability for facilitating andinducing their users to commit infringement.

In this way, innovative technologycompanies are being dragged into court with the threat of possibly having topay large amounts of damages if they are found liable for the individualinfringements of each of their respective users. Given the distributed natureof many of these tools, knowledge of the infringement of individuals issometimes hard or impossible to monitor.

3. Exceptions to Copyright

Copyright laws in the United States do not attempt to prohibit all use or copying of copyrighted materials. [90] Rather, the Copyright Act of 1976 created a “fair use” doctrine that allowssome types of uses of copyrighted material. [91] The fair use doctrine has been used to describe situations in which copyrightedmaterial may be legally used without the authorization or permission of thecopyright holder. [92] Section 107 of the Copyright Act states that fair use of a copyrighted workincludes the creation of copies in order to criticize it or comment on it orfor the purposes of “news reporting, teaching, scholarship, or research.” [93] The factors considered in determining whether a work constitutes fair useincludes 1) the purpose and character of the use, including whether the use is of a commercial nature or is for nonprofit educationalpurposes, 2) the nature of the copyrighted work, 3) theamount and substantiality of the portion used in relation to the copyrightedwork as a whole, and 4) the effect of the use upon the potentialmarket for or value of the copyrighted work. [94] Therefore, while the fair use doctrine does not list specificpermissible uses for each type of work, the Act provides a framework by whichcourts can carve out fair use exceptions. Courts have used the fair usedoctrine to carve out exceptions to copyright infringement in a few file-sharingand online search situations.

B. File-Sharing Legal Background

1. Introduction to File-sharing Case Law

When adjudicating P2P technologies’potential liability for contributory copyright infringement, courts have had toaddress a difficult balance. First, they must promote the interests ofcreativity by upholding a copyright owner’s ability to protect its content. Second, they must promote technological innovation by limiting a newtechnology’s potential liability for infringing copyrights. This comment willdiscuss how courts have attempted to balance these competing concerns.

2. Sony Corp. of Am. v. Universal CityStudios, Inc.

In Sony, Universal Studiossued Sony as the manufacturer of Betamax VCRs. [95] Universal Studios was concerned that VCRs could be used to tape theircopyrighted television shows and allowed users to build up libraries ofcopyrighted material without paying the copyright holders. [96]Universal Studios claimed Sony was liable for contributory infringement becauseSony provided this tool that was capable of direct copyright infringement. [97] Universal Studios sought an injunction that would prevent Sony from selling itsVCR technology. [98] They claimed that the distribution of the VCR combined with the constructiveknowledge that the purchasers of the VCR would use the tool to infringe wasenough to make Sony contributorily liable for their infringement. [99]

In a 5-4 decision, the SupremeCourt applied the “staple article of commerce” doctrine from patent law as aparadigm for contributory copyright infringement. [100] This doctrine required the plaintiff to show 1) direct infringement by theowners of the VCRs, 2) that Sony knew of the direct infringement, and 3) thatSony materially contributed to the infringement. [101] Regarding the second element, the Court held that the distribution of VCRs“does not constitute contributory infringement if the product is widely usedfor legitimate, unobjectionable purposes.” [102] Rather, it need merely be capable of “substantial non-infringing uses.” [103] Therefore, it found that the direct knowledge element could not be met simplythrough the distribution of a product capable of infringement if the productswere also capable of substantial non-infringing uses. [104] To further justify its decision, the Court noted that many of the copyrightholders of licensed broadcasts on free television would not object to havingVCR owners record the shows for later viewing. [105] Going through the analysis of the fair use doctrine in the United StatesCopyright Act the Supreme Court held that this “time shifting” constituted a“fair use” of the copyrighted material. [106] The Court said that holding Sony liable for contributory infringement would“inevitably frustrate the interests of broadcasters in reaching the portion oftheir audience that is available only through time-shifting.” [107]Therefore, it held that since Sony’s VCRs were capable of substantialnon-infringing uses, the mere distribution of the VCRs did not constitutecopyright infringement. [108]

However, the decision in Sonyleft open for interpretation what was required to satisfy the element of directknowledge of infringement. The question of whether mere distribution of aproduct would ever satisfy this element was left open. In A&M Records,Inc. v. Napster, Inc. and In re: Aimster Copyright Litigation, twocircuit courts found different P2P technologies liable for contributoryinfringement. However, they defined the standard for determining liability invery different ways.

3. A&M Records, Inc. v. Napster, Inc.

In Napster, A&M Recordssued the creators of Napster for vicarious liability and contributory copyrightinfringement for creating and distributing its P2P software. [109] Napster argued that it was protected from liability for contributoryinfringement based on the Sony decision. [110] In particular, it argued that the software was capable of substantialnon-infringing use and that its users’ downloading of copyrighted materialconstituted “fair use” similar to the “time shifting” defense employed in Sony. [111] Napster argued that there were a number of “fair uses” of their software including1) sampling, in which users made temporary copies of a song before purchasing,2) space shifting, in which users accessed a song digitally that they alreadyowned on CD, and 3) the permissive distribution of songs authorized by theartists[112]. The issue presented to the Ninth Circuit was whether Napster was liable forcontributory infringement and, if so, whether it had a valid defense of fair use.

The Ninth Circuit started byconfirming that Napster users downloading copyrighted music files were directcopyright infringers. [113] Regarding Napster’s potential liability under a theory of contributoryinfringement, the court applied the “staple article of commerce” doctrinecreated in Sony. [114] The court was forced to analyze the direct knowledge element left ambiguous in Sony. The court held that to find that Napster had direct knowledge of copyrightinfringement A&M Records had to show that Napster 1) had reasonableknowledge of specific infringing conduct, and 2) failed to act to prevent thedistribution of infringing works. [115] The court found that Napster had contributed to the infringement of copyrightsowned by the plaintiffs because it knew of specific incidents of infringementand it failed to filter its index servers to remove references to copyrightedmaterial. [116] Distinguishing Sony,the court said that while VCR manufacturers had no control over how people usedthe technology after they were purchased, Napster could control the infringingbehavior after the software client had been installed by monitoring their indexservers. [117] The courtdiscussed Napster’s centralized software design, with its index servers ofavailable MP3 files, and said that this provided the “site and facilities” forinfringement. [118]

Moreover, the court found thatNapster did not have an adequate fair use defense to the claim of contributorycopyright infringement. [119] It held that sampling did not constitute a fair use because the “sample” wasactually a complete copy of a copyrighted work and would not encourage a futurepurchase. [120] Additionally,the court found that the space-shifting argument was not fair use because inaddition to obtaining the songs in digital format, the software also made thesefiles available to be shared with the rest of the Napster network. [121]

Given that Napster’s architectureplayed a significant role in the court’s decision, it was unclear whether P2Ptechnology might escape liability under a theory of contributory infringementby implementing a decentralized architecture.

4. In re: Aimster Copyright Litigation

Two years after Napster, theSeventh Circuit adopted a narrower view of the protection Sony providesthan the Ninth Circuit. Once again, the court was presented with the issue ofwhat was necessary for determining whether there was direct knowledge ofinfringement as required by the second element of the “staple article ofcommerce doctrine.” [122] Posner held that a balancing test should be deployed that compares thelikelihood of non-infringing uses with the potential extent of illegitimateuses. [123] Posner said thatthe Court in Sony recognized that VCRs were used for both infringing andnon-infringing uses, and that it struck a cost/benefit tradeoff in favor ofSony. [124] In holding thatAimster was liable for contributory infringement, it rejected Aimster’sargument that it was protected because an encryption feature prevented Aimsterfrom knowing what songs were copied by the users of the system. [125] The court held that “willful blindness” can be a basis for finding theknowledge of direct infringement for the purposes of contributory infringementparticularly because Aimster encrypted the files solely so that it might beprotected by the Sony rule rather than as a means of providing some realbenefit to its users. [126]

In discussing his balancing test inthis case, the Seventh Circuit gave weight to the fact that Aimster’s tutorial onlyprovided examples of how to share and download copyrighted files. [127] The court found that this invitation to infringement was exactly what theSupreme Court found missing in Sony. [128] Additionally, the court said that Aimster failed to produce any evidence thatits software had ever been used for a noninfringing use or the frequency ofsuch uses. [129] The court saidthat even if Aimster had shown that its software was used for noninfringinguses, it would also have to show that it would have been disproportionatelycostly to eliminate or substantially reduce the infringing uses. [130] Therefore, since Aimster willfully blinded itself to direct infringement takingplace using its software and failed to take any steps to eliminate infringinguses, the Seventh Circuit found it liable for contributory infringement.

5. MGM v. Grokster

i. Procedural History

In Grokster, the music andmotion picture industries brought an action seeking an injunction against twodistributors of P2P software, Grokster and StreamCast (hereafter “Grokster”),under a theory of contributory infringement. This was the first time the NinthCircuit was presented with a P2P copyright case involving technologies withdecentralized architectures, different from Napster’s centralized index serverapproach. [131] The DistrictCourt granted summary judgment in favor of Grokster, holding that it was notliable for copyright infringement. [132] In upholding summary judgment, the Ninth Circuit followed its analysis in Napsterand applied its interpretation of the Sony doctrine to the knowledgerequirement of contributory infringement. [133] It therefore applied the tests from Sony and Napster and foundthat since Grokster was capable of substantial, commercially viablenoninfringing uses, MGM had the burden of showing that the Grokster hadreasonable direct knowledge of specific infringing conduct. [134] The court held that, given the decentralized nature of Grokster’s product, theprosecution failed to show that Grokster had this direct knowledge. [135] Therefore, the court held that Grokster was simply a distributor of softwareand that they did not materially contribute to the copyright infringement oftheir users. [136] Having a decentralized technical architecture seemed to be the key for thesoftware distributors to be able to escape liability. In addition to eliminatingGrokster’s knowledge of direct infringement, it also prevented Grokster fromhaving the ability to supervise or monitor how their end users were using thetool. [137]

ii. Supreme Court – MajorityOpinion

When Grokster reached the SupremeCourt, many believed the Court would use the case to decide the appropriatetest for determining whether the knowledge element of the staple article ofcommerce doctrine requiring knowledge of direct infringement. The Ninthcircuit’s more technology friendly approach stated that a distributor ofsoftware would only be held liable for contributory infringement if it hadreasonable knowledge of specific infringement and that this knowledge does notexist with a decentralized P2P architecture. On the other hand, the SeventhCircuit’s more copyright holder friendly balancing test gave the court morediscretion in weighing the pros and cons of the technology against thedetriment of the copyright infringement. Which approach would the SupremeCourt adopt?

Rather than answering thisquestion, the majority opinion of the Supreme Court avoided the questionaltogether and decided the case by using another principle of patent law – theinducement theory. [138] Under the inducement theory, the Court held that a software distributor thatpromotes the use of its tool to infringe copyright “as shown by clearexpression or other affirmative steps taken to foster infringement” is liablefor the resulting infringement of their users. [139] Therefore, the Court held that the Ninth Circuit’s granting of summary judgmentwas inappropriate because they did not address whether Grokster tookaffirmative steps to induce its users to commit copyright infringement. TheCourt laid out the test under the inducement theory stating that a company isliable for contributory infringement if they 1) intend to bring aboutinfringement, 2) by distributing a device suitable for infringing use, and 3)actual infringement occurred. [140] The Court reconciled this new rule with the protection Sony providedunder the staple article of commerce doctrine by saying that Sony dealtwith the narrower issue of whether there is “a claim of liability based solelyon distributing a product with alternative lawful and unlawful uses, withknowledge that some users would follow the unlawful course.” [141] Rather, in Grokster the Court was addressing whether there is a claimwhen there is actual intent and affirmative steps taken to encourage copyrightinfringement.

The majority opinion rejected theNinth Circuit’s interpretation of Sony that “whenever a product iscapable of substantial lawful use, the producer can never be heldcontributorily liable for third parties’ infringing use…even when actualpurpose to cause infringing use is shown…unless the distributors had specific knowledgeof infringement at a time at which they contributed to the infringement, andfailed to act upon it.” [142] Rather, the Court held that the rule in Sony does not require courts toignore evidence of the distributor’s intent. [143] The majority opinion then applied this new inducement test to the facts in Grokster. Given that Grokster clearly distributed its product and that there was fairlysubstantial evidence that users used the software to commit actualinfringement, the only element of the inducement rule the Court addressed atlength was whether Grokster took affirmative steps to encourage its users tocommit copyright infringement. The Court noted a number of areas thatsuggested that Grokster had induced infringement including 1) advertising and instructinghow to infringe, 2) targeting former Napster users known to infringe, 3)failing to take steps to prevent infringement, 4) profiting from infringement,and 5) simply distributing a tool capable of infringement. [144] Regarding advertising, the Court said that advertising an infringing use orinstructing how to engage in an infringing use can show an affirmative intentthat the product be used to infringe and can expose the software distributor toliability. [145] The Court heldthat Grokster was not merely a passive recipient of information aboutinfringing use. [146] Rather, the Court found that Grokster wanted recipients to use it to downloadcopyrighted works, and took active steps to encourage infringement. [147] The Court noted that Grokster distributed an electronic newsletter containinglinks to articles promoting its software’s ability to access popularcopyrighted music. [148]

The Court also discussed Grokster’ssimilarity to Napster, a tool used for infringement. [149] Internal e-mails said they wanted to be positioned to capture the flood ofNapster’s users that would be “actively looking for an alternative.” [150] Furthermore, the CTO of Streamcast even said one of their goals was to “get introuble with the law and get sued.” [151] The Court even said the similarity of their product names was evidence of theirintent to profit off of their users’ infringement. [152] Regarding preventing infringement, the court noted that Grokster had made noattempt to use filtering technology to help discourage copyright infringement. [153] According to the Court, this evidence underscored Grokster’s intentions to havetheir users commit copyright infringement. [154] And finally, the Court noted that Grokster’s revenue stream was generated byadvertising. [155] Therefore, given that a large number of Grokster users were using the tool tocommit copyright infringement, the Court found that Grokster was profitingbased on its users infringement. [156] It had translated demand for copyrighted materials into dollars. [157]

Based on all this evidence theCourt held that Grokster’s unlawful objective was “unmistakable.” [158] By contrast, Sony did not try to profit from unlawful taping, it did not intendfor its users to unlawfully tape shows, and Sony did not take affirmative stepsto encourage its users to use VCRs to infringe copyrights. [159]

The copyright holders wished thatthe Court had held that P2P technologies were all tools of infringement thathad not reached the necessary level of substantial non-infringing uses requiredunder Sony. Grokster, on the other hand, wanted the Court to uphold theNinth Circuit’s decision that by nature of the decentralized architecture ithad successfully avoided having the requisite knowledge of infringing activityrequired for contributory liability. By deciding the case using anintent-based inducement theory of contributory infringement, the majorityopinion avoided this question of when a technology fails to exhibit sufficientpotential for lawful uses and therefore imputes on the distributor the directknowledge of infringement required under the staple article of commercedoctrine. Therefore, the message from the majority seems to be that under theinducement theory you must show by the defendant’s own actions and statementsthat his unlawful purpose disqualifies him from claiming the protection under Sony. That is, clear evidence of intentional inducement trumps, even in situationswhere the technology may have substantial non-infringing uses. However, themajority declined to address the conflict about when knowledge of infringementmay be imputed to a P2P software distributor saying that it would “leavefurther consideration of the Sony rule for a day when that may berequired.” [160] While all ninejustices signed on to the majority opinion, Ginsburg and Breyer both wroteseparate concurrences in which they both touched on this issue. In hisconcurrence, Ginsburg effectively argued for the Seventh Circuit’s approach todetermining the knowledge component that would require the court’s to employ abalancing test and would give them more discretion to protect copyrightholders. [161] In hisconcurrence, Breyer basically attacked Ginsburg’s opinion as not doing enoughto protect technological innovation. He argued for something closer to theNinth Circuit’s narrower test. [162]

iii. Supreme Court – Ginsburg’sConcurrence

In Ginsburg’s concurrence, sheargued that the inducement theory could lead to liability and therefore was anadditional tool to protect copyright holders. [163] However, the primary purpose of the concurrence was to say that the NinthCircuit erred in granting summary judgment to the claim of contributorycopyright infringement based on product distribution alone. [164] In addition to addressing whether Grokster was liable under the new inducementtheory, Ginsburg believed the court should reconsider whether summary judgmentwas appropriate for contributory infringement under the traditional test ofsubstantial non-infringing uses under Sony. [165] Under Ginsburg’s approach, liability could be predicated on inducinginfringement or simply by distributing a product used to infringe copyrights ifthe product is not capable of substantial, commercially significantnon-infringing uses. [166] She argued that Grokster had not made a sufficient showing of substantial non-infringinguse to qualify for summary judgment, saying that there was only “anecdotalevidence” of the P2P technologies being used for non-infringing uses. [167] Therefore, Ginsburg argued for a balancing test more in line with Posner’srecommended balancing test in the Seventh Circuit’s Aimster decision. Ginsburg stated that determining whether there were “substantial”non-infringing uses could only be done by comparing the number or percentage ofinfringing use to the non-infringing uses. [168]

iv. Supreme Court – Breyer’sConcurrence

Breyer’s concurrence, on the otherhand, while agreeing with the majority about the potential liability under theinducement theory, disagreed vehemently with Ginsburg’s concurrence saying thatthe court should reconsider granting summary judgment for contributoryinfringement under the rule in Sony. [169] Therefore, Breyer thought the Ninth Circuit got it right when it grantedsummary judgment on the issue of whether Grokster’s product was capable ofsubstantial non-infringing uses. Breyer saw the deeper issue in this casebeing whether the Court should modify the Sony standard to broaden the scope ofcontributory copyright infringement claims as Ginsburg suggests. [170] Breyer argued that MGM failed to demonstrate the need to modify Sony andthat modifying Sony would impose significant additional risks ontechnological innovation.[171]

Breyer argued that Grokster passesthe Sony test because it is “capable of” substantial non-infringinguses. [172] Examples ofthese non-infringing uses included sharing 1) authorized copies of music fromartists like Wilco, Pearl Jam, Dave Matthews, and John Mayer, 2) freeelectronic books and other works from online publishers, 3) public domain andauthorized software such as Winzip, and 4) licensed music videos and televisionand movies that are distributed via digital video packaging with the permissionof the copyright holder. [173] Breyer said it was important that Sony asked whether the product was“capable of” substantial non-infringing uses. [174] The language, he argued, suggested that a court could look at the reasonableprospect of expanded legitimate users over time. [175] Therefore, the tool does not even require substantial non-infringing uses inthe short term if it appears that there is a reasonable prospect for there tobe substantial non-infringing uses in the future. [176] Therefore, Breyer argued that as more and more non-copyrighted material becomesavailable in digital form, it is reasonable to assume that lawful P2P sharingwill become more prevalent. [177]

He argued that these non-infringinguses would be likely to increase for sharing things like 1) general researchinformation, 2) public domain films, 3) historical recordings and digitaleducation materials, 4) personal photos, 5) licensed movies and music files,and 6) news broadcasts. [178] For Breyer, determining whether to modify the rule from Sony, as heunderstands it, requires assessing whether the modification would weaken theprotection of technology so much as to outweigh the new copyright-relatedbenefits. [179] He argued thatthe rule in Sony was very clear and that it rightly recognized that thegoal is not to discourage or control the emergence of new technologies,particularly ones like P2P technology that help disseminate information and ideasmore efficiently and broadly. [180] He argued that the “capable of” language in Sony is forward looking anddoesn’t confine the analysis of liability to a static interpretation of theproduct’s existing uses. [181] Therefore, it protects technologies that may have undeveloped future markets. [182]

Moreover, he believes the Sonyrule is good because it recognizes that judges don’t have specialized technicalability to answer questions about the feasibility or commercial viability ofdifferent products. [183] He used the example of filtering technology which Grokster argued would be verydifficult and inefficient to create and implement and which MGM claimed wouldbe very easy to do. [184] Under Sony, Breyer believes the judge wouldn’t necessarily have todecide these types of technical questions. [185] Moreover, Breyer argues that using a balancing test like the one proposed inGinsberg’s concurrence would significantly weaken the law’s ability to protectnew technology. [186] To employ Ginsberg’s balancing test, defendants would potentially have toprovide detailed evidence of their business plans, profitability estimates,projected feature development cycles and more in order to prove that thebenefits of eventual potential lawful uses of the technology outweigh any shortterm issues of copyright infringement. [187] Inventors and entrepreneurs would be subject to potential lawsuits andextensive trials whenever they create and distribute any sort of technologythat could be used to infringe copyrights. [188] These inventors would have no way to predict how the courts would weigh thevalues of infringing versus non-infringing uses. [189] This risk and uncertainty, according to Breyer, would slow down the developmentof new technologies.[190]

Finally, Breyer argued that inorder to decide to modify the Sony rule to better protect copyrightholders from the infringements taking place in P2P networks, MGM had to showthat the benefits to better protection of copyrights clearly outweighed thisweakening of technological innovation. [191] Breyer found that the evidence did not make out a very strong case for changingthe rule in Sony. [192] He argued that it was unclear how much revenue to the copyright holders haddecreased as a result of the infringement. [193] Furthermore, he argued that part of the role of copyright protection is toensure that there are adequate incentives to produce creative works and thatthere was good reason to believe that any decrease in production of creativeworks under the Sony rule was not substantial. [194] In particular, revenue flowing to musical artists from performances and othersources would remain stable even if the entire CD market was replaced by P2Pdistribution. [195] He went so far as to say that it would be “silly” to think that music, “acultural form without which no human society has existed”, will go away becauseof P2P technology. [196] In fact, Breyer quoted surveys saying that 70% of musicians thought that P2Ptechnology was a minor threat to no threat at all to creative industries. [197]

Moreover, Breyer argued thatcopyright holders had other avenues they could pursue to reduce infringement oftheir works. [198]For example, they could go after the direct infringers of the copyrights. [199] Breyer quoted statistics showing that when copyright holders pursued thesetypes of lawsuits they had a significant deterring effect.[200] Moreover, the copyright holders could be proactive in developing new technologythat would help reduce unlawful infringement. [201] For example, new technologies like “digital watermarking” and “digital fingerprinting”could be used to encode information into files that could help prevent theirmisuse. [202] Additionally,Breyer quoted evidence that people were willing to pay for P2P style servicesthat sold music over the internet. [203] He argued that many consumers that initially appreciated the convenience of aservice like Grokster are now migrating to lawful services in which the cost isworth the benefit of additional convenience and flexibility without the risksof engaging in unlawful infringement. [204] And finally, Breyer argued that courts are less well suited than Congress toattempting to balance the competing interest that inevitably arise with new,emerging technologies. [205]

C. Search Engine Legal Background

1. Introduction to Legal History

After search engines began to beused prevalently to organize and help find the world’s digital information,legal disputes soon followed. [206] Many lawsuits had to do with trademark issues. [207] For example, search engines were dragged to court by companies that werefrustrated that the search engine sold advertising rights to one of theircompetitors based on search criteria that included that company’s trademarks. [208] Moreover, search engine caching and the indexing of copyrighted web contenthave led to the question of what constitutes fair reuse of copyrighted webcontent.

2. Kelly v. Arriba Soft Corp.

After Napster, the NinthCircuit was forced to address a different type of fair use. In this case,Kelly was a photographer who made his copyrighted pictures available over theInternet. [209] Arriba Soft tookKelly’s pictures and made thumbnails, small low quality versions of digitalpictures, which they made available through their search engine. [210] Arriba’s database did not store the full picture but only the thumbnail. [211] Clicking on the thumbnail would take the user directly to Kelly’s website, inan Arriba Soft web frame, where the full picture was made available. [212] Kelly sued Arriba for direct copyright infringement. [213] The lower court found that Arriba Soft violated the copyrights and did not havea valid fair use defense for providing the thumbnails from within their searchengine. [214]

On appeal, the Ninth Circuitreversed holding that the thumbnails did constitute fair use. [215] Therefore, Arriba Soft was allowed to display copies of images (acceptableresolution was not decided) in their search results. In analyzing the fair usedoctrine, the Ninth Circuit used Section 107 of the United States Copyright Actas its guide. [216] Regarding “purpose and character of the use,” the court held that the use wascommercial, but not of the same character as the original work. [217] The images were not being sold and weighed in Arriba’s favor because there was“public benefit” to the search engine. [218] Regarding the “nature of the copyrighted work,” the court held that the natureof pictures as published creative work was a factor that worked slightly infavor of Kelly.[219] Regarding the “amount and substantiality” of the portion of the work used, thecourt said that it was necessary for Arriba to copy the entire image in orderto allow their users to determine whether they wanted to pursue the link andobtain more information.[220] Therefore, this factor did not weigh heavily in either parties favor. [221] Regarding the “effect of the use upon the potential market for or value of thecopyrighted work” the Ninth Circuit found that the transformation of the imagesfor use in the search engine would not have an adverse effect on Kelly’s marketfor his images. [222] Rather, it found that the thumbnails would guide larger number of users toKelly’s work rather than away from it. [223] The thumbnails themselves would not detract from this market because the sizeand quality would make them unattractive as a substitute for the originalimages.[224]

Concluding that two of the fair usefactors weighed heavily in Arriba’s favor and only one weakly favored Kelly’sposition, the Ninth Circuit carved out a fair use exception for usingthumbnails of copyrighted images in search engine databases. [225]

IV. ANALYSIS

A. An Assessment of BitTorrent’s Potential Liability

1. BitTorrent’s PotentialLiability Under an Inducement Theory

In Grokster, the Court laidout the test under the inducement theory stating that a company is liable forcontributory infringement if they 1) intend to bring about infringement, 2) bydistributing a device suitable for infringing use, and 3) actual infringementoccurred. Given what some have called the "Miss Manners" rule, itseems that to avoid liability under an inducement theory companies must simplyavoid marketing their products as ones capable of direct copyrightinfringement. Companies will begin setting up processes to ensure theirinternal communications and externally facing documentation could not beinterpreted as having wrongful intent in distributing their product. Therefore, this rule makes it increasingly difficult to tell the differencebetween companies that truly don’t want infringement to take place and thosethat do want it, but are able to sufficiently hide their intentions. So howwill BitTorrent fair under this new rule of secondary liability? BitTorrentshould be protected. As discussed in the Technical Background, BitTorrent is aP2P technology that is very good at sharing large files. One study claims thatin 2005 over 35% of all Internet traffic was taking place over the BitTorrentnetwork.

BitTorrent is different fromNapster, Aimster and Grokster partially because BitTorrent refers to both 1)the communication protocol and 2) a default application employing the protocolto share files. It is clear that the the BitTorrent company should not beliable, under any theory, for the mere creation of a new communicationprotocol. However, the question is whether they would be liable under aninducement theory for the distribution of a P2P file-sharing application thatemploys that protocol. While there is no doubt that BitTorrent is anapplication that is suitable for infringing copyrights and that actualinfringement on the part of its users occurs, a court should not find that theyhave the requisite intent needed to establish secondary liability under theinducement theory.

BitTorrent’s original customer wasetree, a community of pepole who distribute live concert recordings for freeonline, with the permission of the bands. Bram Cohen, BitTorrent’s founder hassaid that he developed BitTorrent based on the needs of his friends to legallyshare and download these types of shows. Additionally, there have been othersignificant commercial uses of the technology. The protocol has been used todistribute patches for Blizzard Entertainment’s World of Warcraft game and manysoftware companies use it to distribute builds among their technology teams. In Grokster, the companies distributing file-sharing applications were foundliable under an inducement theory because they 1) specifically sought tocapture Napster’s market of known copyright infringers, 2) they activelyinduced their users to infringe through their instruction manuals and customersupport networks, 3) failed to filter out copyrighted materials, and 4)profited from the infringement through advertising revenue. First, BitTorrenthas not tried to capture a pre-existing group of copyright infringers from adifferent technology. Rather, from the very beginning Bram Cohen has made itclear that his intent is to serve a legitimate market of providing tools forlegally sharing and downloading concerts. Second, BitTorrent does not seem toinduce infringement through their documentation. Their FAQ section specificallyoutlines their goals and purpose. Even a very early version of the BitTorrentFrequently Asked Question (FAQ) list said that BitTorrent was developed forlegal trading of these types of shows. Moreover, when the company was testingthe initial versions of the technology, they used freely available,non-copyrighted video content. By contrast, the examples in Grokster’sdocumentation all involved infringing content.

Bram Cohen did write, in hisTechnological Activist’s Agenda, that he "build[s] systems to disseminateinformation, commit digital piracy", etc. However, while this statementwas available from the BitTorrent website for about two years from July2001-July 2003, a lawsuit under active inducement should not read much intothis statement. Sony also created a system that could be used to disseminateinformation and commit digital piracy. Simply saying that the system can beused to commit digital piracy is not the same thing as actively inducing othersto use the product in this way. Third, while BitTorrent does not provide toolsfor filtering out copyrighted material, this should not weigh heavily againstBitTorrent. Courts and judges are not in the best position to judge thequality and efficiency of a particular product design. They are notwell-positioned to be making cost-benefit analyses of different feature sets orproduct designs. Therefore, while the Grokster court did mention thisas evidence of Grokster’s intent, the court’s are likely to limit the impact ofthese types of product design questions on their decisions of finding intent. Fourth, while Grokster and Streamcast were profiting from infringement throughtranslating demand for copyrighted materials into advertising dollars,BitTorrent is entirely free and does not include advertising in its defaultclient application. That being said, BitTorrent recently added a search engineto their website that does include paid advertisements. This will probably bethe biggest hurdle for BitTorrent to overcome in proving they were not inducingor encouraging infringement. In the same way the Court found that Grokster wasconverting demand for copyright infringement into advertising dollars, a courtcould similarly find that BitTorrent’s new advertising based search capabilitiesserve the same purpose. Moreover, BitTorrent’s recent move to a decentralized,trackerless technology design may make it more difficult to escape liability. While the court never specifically faulted Grokster for developing adecentralized technology in an attempt to escape liability, a court could findthat movement in this direction was evidence of intent to have its usersinfringe. In the same way Grokster tried to avoid liability with itsdecentralized use of Supernodes to avoid centralized index files there is anargument that BitTorrent’s move to trackerless torrents could be seen asevidence of intent to have its users infringe.

However, the overwhelming evidenceis that BitTorrent has not "shown by clear expression or other affirmativesteps" that they intend for their users to infringe copyright. Theconsistency by which BitTorrent’s owner, Bram Cohen, has stated BitTorrent’slegal purpose and target markets and the lack of documentation or tools forhelping users use the technology to commit infringement should be enough toprotect BitTorrent from liability under an inducement theory. Nevertheless,the new inducement theory creates less predictability for developers andtechnology companies. It increases the cost of the entire legal process bymaking discovery more expensive as it requires more investigation to discoverand identify a product company’s true intent.

2. BitTorrent’s PotentialLiability Under Contributory Copyright Infringement

While both Napster and Aimster werefound liable for contributory copyright infringement, the circuit courts hadtwo different tests that emerged to determine whether the defendant had therequisite knowledge required by the second element of the "staple articleof commerce" doctrine adopted from patent law in Sony. In theNinth Circuit, drawing heavily on Napster’s centralized architecture the brightline test required the prosecution to show that 1) the defendant had reasonableknowledge of specific infringing conduct, and 2) the defendant failed to act toprevent the distribution of infringing works. In the Seventh Circuit, thecourt would apply Posner’s balancing test that would weigh the evidenceincluding 1) percentage of noninfringing to infringing use, 2) whether thedefendant "invited" infringement (e.g. in its documentation andtutorials), and 3) whether defendant took steps to eliminate infringing uses. Whilethe Supreme Court did not address this issue directly in Grokster,Ginsberg and Breyer’s two concurring opinions discussed very different views onwhether a technology company could be liable based solely on the distributionof a product capable of infringing use. Breyer’s rule is better at promotingtechnological innovation and protecting the interests of future, undeveloped marketsthat could benefit from emerging technologies.

i. Liabilityunder Ginsberg’s Balancing Test

Under Ginsburg, BitTorrent wouldhave a harder time escaping liability. Ginsberg more closely followed theSeventh Circuit balancing test saying that a technology company can be liablefor contributory infringement based on distributing a product capable ofinfringement alone. A company that distributes a product that is not capableof substantial, commercially significant, non-infringing uses would be liablefor the copyright infringement of their users. Therefore, BitTorrent wouldhave to show that its technology had substantial, non-infringing use that werenot simply the "anecdotal" evidence she saw in Grokster. Sherecommends the balancing test of looking at the percentage of infringing andnon-infringing uses in order to determine whether they are"substantial." Therefore, this test would essentially make the testfor secondary liability based on contributory copyright infringement a predominantuse test. If the software is predominantly used for infringing uses, it isliable for contributory infringement. If the software is predominanly used fornon-infringing uses, it escapes liability. BitTorrent would probably fail thistest because currently, a large portion of its users use the technology todownload copyrighted movies and television shows. It would be more difficultfor Ginsberg to simply dismiss the evidence of non-infringing BitTorrent usesas being "anecdotal." For instance, unlike Grokster, BitTorrent wascreated with a lawful, non-infringing use in mind - sharing and downloadingconcerts with the permission of the bands. Moreover, there are a number ofexamples of commercial uses of BitTorrent to distribute patches for video gamesand software, as well as a tool for distributing large applications like Linuxor other operating systems. Nevertheless, when comparing strict percentagesthe infringing use very likely dwarfs the non-infringing use.

This is troubling given the vastpotential for the technology. Some people consider BitTorrent to be theworld’s largest TiVo. They believe it is a technology that willrevolutionalize the distribution and production of content as more contentproviders model their creative works for online audiences. One example was agraduate computer science student who enjoyed Outfoxed, a documentary thatcriticized Fox News. The student convinced the film’s producer to allow him tohost the file online as a .torrent file. 1,500 people downloaded it in thefirst two months, effectively turning this computer science student into whathe called "a movie producer." "If I had my own content, I’d bea TV station." A similar example was when Jon Stewart, host of The DailyShow, appeared on CNN’s Crossfire and verbally insulted the hosts calling thempolitical puppets. The content was made available as a .torrent and within asingle day 4.000 servers were swapping the clip. Over two million peoplestreamed it from a website over the next few weeks while only ~850,000 peoplesaw the live broadcast. More and more people will begin watching TV and filmcontent online. However, for these changes to take place the technologyfacilitating these revolutions must be protected.

If the correct interpretation ofGinsberg’s balancing test is not simply a predominant use test, but ratherrequires the more careful balancing of potential future markets, developmentplans to reduce infringement, etc., the test is still inefficient. Under thisunderstanding of Ginsberg’s approach balancing test, BitTorrent would have toprovide detailed evidence of its business plans, future market opportunities,projected feature development cycles and potential revenue streams in order toprovide the court with the information needed to perform the balancing test. It must show convincingly that the evidence of its commercial uses are morethan the examples Ginsberg considered "anecdotal" in Grokster. While this will provide BitTorrent more leeway in proving its potential valuethan under a strict predominant use test, the Court would still be likely tofind BitTorrent liable for contributory infringement.

ii. Under Breyer’s “Capable of”Substantial Non-Infringing Uses Approach

BitTorrent would almost certainlybe excused under Justice Breyer’s approach. Under Breyer’s understanding ofSony, granting Grokster summary judgment based on a theory of contributoryinfringement for distribution of a product capable of infringing use wasappropriate. He argued that Sony passed this test because Grokster was"capable of" substantial, non-infringing uses and that this was not abalancing test as Ginsberg would like. He argued that the ability to useGrokster to share authorized music, free electronic books, and licensed videofiles were all evidence of non-infringing uses that could reasonably beexpected to expand over time. This same argument holds true for BitTorrent. Therefore, under Breyer’s rule BitTorrent would be free from liability undercontributory infringement because as more and more non-copyrighted or licensedmaterial became available in digital form, it would be reasonable to assumethat lawful P2P file-sharing would become more prevalent.

This is a better rule thanGinsberg’s balancing test. Breyer’s reading of Sony does a better job ofprotecting the interests of users and creators of P2P software while avoidingthe problems associated with implementing Ginsberg’s balancing test. Ginsberg’s balancing test would lead to inconsistent rulings as differentjudges weighed the benefits and costs differently based on their subjectivevaluations. Moreover, Breyer’s understanding of the Sony rule better protectstechnologies that may have undeveloped future markets. The nature oftechnology is that its innovation forces companies and individuals toreconsider the best, most efficient means of executing on their businessplans. Technologists should be encouraged to invent without the fear of beingdragged into court for the mere creation and distribution of a tool that mightinitially be used primarily to infringe copyright. Second, this rule is moreeffective because it recognizes the limitations judges have in answeringtechnical questions about the commercial viability of different products orproduct features. Breyer’s rule recognizes the important distinction thattechnology does not commit infringement, people do. Rules of contributoryinfringement must protect the interests of technologists creating andinnovating in order to find new solutions to problems. The burden should be oncopyright holders to develop appropriate means of protecting their rights. AsBreyer’s points out, if the movie and television industries are frustrated atthe amount of their content being distributed over BitTorrent they areperfectly capable of creating new technologies like “digital watermarking” or“digital fingerprinting” that could be used to prevent the misuse of theircontent.

Moreover, if they did develop these types oftechnologies, Breyer’s rule would protect them against tools that would decodetheir encryptions in order to remove their digital rights managementtechnologies. Not only is this type of obstruction covered in the DigitalRights Management Act, but it is similar to descramblers and other technologiesthat courts have held do not have “substantial” non-infringing uses.

B. An Assessment of Google’s Potential Liability forDirect Copyright Infringement

1. Google Book Search

i. Direct Copyright Infringement

As discussed in the Technical Background,Google Book Search has two key facets: the Print Publisher Program and thePrint Library Program. The PrintPublisher Program does not present any copyright issues because under thisprogram Google obtains authorization from copyright holders to scan the text ofcopyrighted books into Google’s search engine. However, under the Print LibraryProject, Google has partnered with major libraries to scan books without theexpress permission of the original publishing companies. The scanning ofcopyrighted books and creation of multiple digital copies (one for thelibraries and one for Google’s search index) constitutes direct copyrightinfringement.

There are three aspects of theproject that could present copyright issues: 1) the digital copy Google scansfor use in its search database, 2) the display of a portion of that scannedtext to users in search results, and 3) the creation of a digital copy of thebooks for the libraries. It is pretty clear that creating a digital copy of anunauthorized book for use in its search database is a copyright violation,unless a fair use exception applies. Similarly, the display of a portion ofthat scanned text to users in search results constitutes a fair use exceptionunless the fair use exception applies. The creation of a digital copy for useat the libraries is not as clear cut. If the libraries, in purchasing thebooks for their library, purchased the rights to create digital copies forarchival purposes then contracting out to have an organization like Googlecreate these copies for them does not seem to be a copyright violation. However, if the libraries do not have the right to make a digital copy fromtheir hard copies, then this too constitutes a copyright violation. However,it is more likely that this would be seen as a violation on the part of thelibraries rather than Google. Therefore, for the fair use analysis, thiscomment will look only at the first two examples of direct copyrightinfringement. Therefore, the primary legal question is whether Google’s PrintLibrary Project, and in particular the creation of unauthorized digital copiesof copyrighted books for use in its search engine, constitutes a fair use ofthe copyrighted work and therefore escapes liability under the fair useexception outlined in Section 107 of the United States Copyright Act.

ii. Fair Use Defense

In order for Google’s Print LibraryProject to escape liability under the fair use doctrine, an analysis of theelements of the fair use test must be made. In particular, when decidingwhether a fair use exception exists, the courts must take into account 1) thepurpose and character of the use, including whether the useis of a commercial nature or is for nonprofit educational purposes, 2) thenature of the copyrighted work, 3) the amount and substantiality of the portionused in relation to the copyrighted work as a whole, and 4) the effect of theuse upon the potential market for or value of the copyrighted work.

Regarding thepurpose and character of Google’s use, the court in Arriba asked whetherthe use was commercial in nature. It found that since Arriba’s use of Kelly’simages was not “highly exploitive”, the commercial nature of the search engineonly weighed slightly in favor of Kelly. Regarding the display of sentences ofsurrounding text to end users in search results, Google is attempting to profitfrom providing these portions of copyrighted material. First, Google willeventually use advertising on these pages. Therefore, similar to Groksterin which the Supreme Court found that the P2P file-sharing companies had founda way to convert users copyright infringement into dollars, Google has found away to profit off of the unauthorized digital reproduction of copyrightedwork.

Nevertheless,Google’s Book Search is a tool that will bring the existence of relevant booksto the attention of potential users. Therefore, despite the purpose of itsinfringement being commercial in nature Google has a fairly strong publicpolicy argument suggesting that its purpose is for the general public good andwill not act to supplant the revenue streams of the copyright holdersthemselves. Rather than being of the character or purpose that would directlycompete with sales of the books themselves, Google Book Search acts to increaseexposure to books and increase revenue for publishers. This is similar to Arribain which the court held that Arriba’s use of Kelly’s images promote the goalsof the Copyright Act, and the fair use exception in particular, by benefitingthe public by making information more readily available. Moreover, it heldthat Arriba’s use of thumbnail versions of Kelly’s images did not supplant theneed for the originals because there was not enough value that could be gleanedby the low quality thumbnails themselves. Similarly, most users of Google BookSearch will not use this tool as an alternative to buying the book itself. Thevalue they receive from getting a few sentences of material, in most cases,will not be enough to supplant book sales in the same way that low quality thumbnailpictures will not supplant the need for the originals. Therefore, despiteGoogle benefiting by a larger stream of advertising revenue through its use ofthese copyrighted works, this would be counterbalanced by the public policybenefits of having this content readily available to the general public and thefact that the model would be more likely to drive revenue for publishers ratherthan supplant their existing revenue streams. Therefore, it is unlikely that thepurpose and character of Google’s use, though commercial in nature, would weighheavily against Google.

Second, thenature of the copyrighted work must be considered. Here, the court in Arribasaid that one of the core aspects of this element of the fair use doctrine wasto ensure the protection of works that are creative in nature rather than “morefact-based works.” According to Jonathon Band, the “vast majority” of theworks scanned under the Print Library Project will be non-fiction “fact-basedworks” rather than creative fictional works. Moreover, according to the courtin Arriba, published works are more protected by the fair use doctrine becausethe “artist’s expression has already occurred.” Therefore, given that the vastmajority of the Print Library Project will be to scan published non-fictionbooks that are more fact-based than fiction, this element will only weighslightly against Google.

Third, theamount of the copyrighted work as a whole that is used must be considered. Inmany cases, quotes are used from copyrighted text. This constitutes fair usepartially because the portion of the work that is used is not a substantialpart of the whole. In the case of Google, the two separate copyrightconsiderations require slightly different analysis. The full reproduction ofcopyrighted text and the storing of entire digital copies of these texts intheir search databases seems to weigh heavily against Google. However, thedistribution of a few sentences of that text that match users search criteriaseems to be of the same nature as quoting copyrighted work that is generallyaccepted as fair use. In Arriba, a full copy of the copyrighted work ismade available, albeit in lower quality. The court held that “if the secondaryuser only copies as much as is necessary for his or her intended use” that thisfactor of the fair use doctrine would not weigh heavily against them. It couldbe argued that in order to fulfill the purpose of providing instant, searchableaccess to all of the world’s books, Google can only achieve this with a full,digital copy of the text. Searches restricted to portions of text would not beas reliable, accurate and relevant as full text-based searches. Nevertheless,Google will have difficulty arguing that the full reproduction of copyrightedtext in their search database is not a “substantial portion” of the copyrightedwork “as a whole.” Therefore, despite the public policy benefits, this factoris also likely to weigh against Google.

Regarding thefourth factor, there are two interesting arguments. Google can argue that, ifanything, its program increases the potential markets for thecopyrighted works. The availability of full text-based search could increasedemand for some books which could cause some users to buy books they otherwisewould not have purchased. While the display of a few sentences of text aroundthe users search criteria probably would not be enough to reduce demand forpurchasing the entire book, this becomes a slippery slope. Google hasincentive to provide as much of the copyrighted work as would be authorizedunder the law. The publishers would be incented to allow Google to providejust the right amount of information to pique the user’s curiosity, make clearthat the content is relevant to the users needs, and encourages the purchase ofhard copies. However, even if providing a few sentences of content throughGoogle Book Search would not significantly displace sales of new books, it doesimpede a publisher’s ability to license new digital copies of its content tolibraries or to search engine providers like Google. For example, in scanningbooks from libraries Google is essentially paying for the rights to fullyreproduce copyrighted works in its search technology with a digital copy thelibraries can use for its archival purposes. Therefore, Google’s method ofpayment is, in effect, an illegal reproduction of a copyrighted work.

Moreover,Google has an adequate alternative. Google could pay publishers for the rightto provide this search feature on their website. Moreover, if Google were notcreating these types of contracts with libraries, these publishers would havethe ability to sell digital copies of their content to libraries for archivalpurposes themselves. Therefore, by pursuing the unauthorized reproduction ofcopyrighted work and giving a copy of those works to the libraries providingaccess to books, Google cuts off two possible modes of revenue to the copyrightholders themselves. Therefore, even if Google’s technology drives new revenueby making books more accessible online, it is not clear that this increase inrevenue will outweigh the loss in potential revenue to publishers who wouldotherwise be able to charge libraries for digital copies and license thecontent to search providers for inclusion into tools like Google Book Search.

If copyright is to mean anything,companies like Google cannot be authorized to make digital copies of entireworks that they have not purchased for commercial gain. Therefore, given thatall four of the fair use factors seem to weigh against Google, it is unlikelythat Google Book Search will be protected from liability under copyright lawunder a “fair use” exception. Rather than circumventing the publishers bygoing directly to libraries for access to copyrighted materials, Google shouldnegotiate with publishers to license access to their works for inclusion in itsGoogle Book Search feature.

CONCLUSION

Given the current state ofcopyright law in the United States, BitTorrent should escape secondaryliability for copyright violations of its users based on an inducement theory. As a new P2P file-sharing technology, its decentralized architecture and itsoriginal intentions to serve a valid, lawful market of providing access todigital recordings of concerts, there does not appear to be the “affirmativesteps” required to induce its users to commit infringement. However, under atheory of contributory copyright infringement, the legal conclusions are a bitmurkier. BitTorrent’s liability is likely to hinge on whether the court adoptsthe type of balancing test promoted by Ginsberg’s concurrence in Groksteror whether the Court adopts the more technology friendly solution in Breyer’sconcurrence arguing that liability should only exist if the product is notcapable of substantial non-infringing uses. Under Breyer’s rule, BitTorrentwould likely be free from liability whereas under Ginsberg’s rule BitTorrentprobably would be liable for the infringing acts of its users.

Google Book Search, on the otherhand, probably does not constitute a fair use of copyrighted material. Bycircumventing the copyright holders and scanning content directly fromlibraries it is making full, digital reproductions of copyrighted works andprofiting from those digital reproductions. Furthermore, it is cutting off twopotential revenue streams to publishers including the ability to license accessto these digital copies to search engines like Google as well as cutting offthe market for selling digital copies of their content to libraries forarchival and search purposes. An analysis of the factors for consideration ina fair use analysis all weigh in favor of the copyright holders, suggestingthat the fair use exception does not apply and that Google is liable.

Both of these cases represent thefailing of copyright law in the digital age. The fact that a technologycompany like BitTorrent cannot say for certain whether it would be liable undertheories of secondary liability for the copyright violations of its usersrepresents a failure of the judicial system to present clear rules withpredictable results. Moreover, publishers’ ability to hold up the process ofmaking books available on the Internet works against the interests of consumersand the general marketplace and represents the inflexibility of copyright lawto adapt to the changing requirements of the Information Economy.

Posted in Law and Technology | No Comments »

INFORMATIONPRIVACY AND THE MODERN GLOBAL FIRM

January 12th, 2007

Posted in Business Process Management | No Comments »

Substantive Due Process: The Right to Privacy

May 8th, 2006

Quick Background on Substantive Due Process
The concept of substantive due process was invented by the Taney Court. It represents a broad, probably inaccurate, interpretation of the due process clause of the Fifth Amendment which states that, no person shall “be deprived of life, liberty, or property, without due process of law.” Historically, “due process of law” protected only procedural rights, such as the right to a fair trial, advance notice of charges being brought against you, conviction by a jury of one’s peers, etc. Therefore, the DP clause was primarily intended to protect American’s from procedural abuses, often by judges. Ironically, the SC has creatively interpreted this clause to protect certain substantive rights not outlined in the constitution. Therefore, a clause that was meant to protect American’s from judicial abuse is now being used as a growing source of unintended judicial power.

Under the judicially created concept of substantive due process, what is deemed constitutional depends on whether the nature of the interest is 1) economic or social, or 2) fundamental person rights.

Economic or Social Rights
In the first part of the 20th century, the SC reviewed the substance of legislation and used the DP clause to invalidate economic and social regulations. The basic rationale was that the legislation unreasonably interfered with liberty, property and freedom of contract. However, these were made by the personal judgments of the Justices about whether the means used were reasonably related to a legitimate end.

For example, in Lochner v. New York (1905), the SC held that a law limiting the number of hours bakers could work interfered with their freedom to contract and was therefore a violation of DP under the 14th Amend. The SC basically felt they had an obligation to protect the free-market system by reading DP as a broad protection of liberty, including the freedom to K and other fundamental rights. Therefore, in the Lochner era, DP was broadly interpreted to protect economic interests as well as non-economic interests like education and marriage.

This is a little frightening. By using substantive due process, the Justices seemed to be stretching the concept of “liberty” and “property” further than the founders would have intended. They were able to read their personal economic predilections directly into the Constitution. Indeed, if you take a sufficiently broad interpretation of these terms, there would be almost no law that the legislative branch could create that the SC could not selectively, based on their personal judgments, strike down.

It seems the SC itself considered this to have gone to far. As a result, the modern approach in economic and social regulations has been to defer to legislative judgments, providing only a minimal level of judicial review. United States v. Carolene Products (holding that the SC will not weigh the wisdom of the legislation or substitute its own judgment for that of the legislative body); West Coast Hotel Co. v. Parrish (1937) (SC upheld a state law establishing a minimum wage for women saying that the health of women was a legitimate end and the creation of a minimum wage was neither arbitrary nor capricious); Williamson v. Lee Optical (1955) (holding that the SC cannot strike down state laws dealing with economics just because they are unwise. Rather, people should protect against legislative abuse by voting). As a result of this line of cases, the SC has not struck down any economic regulation by state legislation as a violation of substantive due process since 1937 and Lochner’s overreaching seemed controlled.

Fundamental Person Rights
However, the SC did not stop there. While the SC was abusing its power to stretch the term “liberty” in the economic and social policy arena, it also began leveraging substantive due process to protect what it called “fundamental rights” that were “implicit in ordered liberty.” While there is no clear way of identifying which rights are fundamental other than looking into psyches of each individual Justice, some include right to refuse medical treatment, the right to travel and the right to privacy (e.g. abortion, sexual activity, parental rights).

Meyer v. Nebraska (1923) and Pierce v. Society of Sisters (1925), are considered to be the first cases where the SC used substantive due process to protect civil liberties. In Meyers, a teacher was convicted of teaching German to a student in violation of a statute prohibiting teaching foreign languages to students prior to 8th grade. The SC did not draw any clear line about how much educational regulation was too much. In Pierce, the SC even allowed the 14th Amend to apply to corporations, rather than just individuals. Over the next half century, a broadened list of liberties and rights became protected including the right to marry, the right to privacy, etc.

What about Privacy rights?
Although most American’s seem to feel an inherent right to a certain extent of privacy, this right is not guaranteed in the text of the Constitution. When the SC initially upheld this right, it did so using a few different theories. First, sometimes they considered it a right protected by the “penumbra” of the Bill of Rights, reading “privacy” between the lines of freedom of the press, search and seizure, etc. Other Justices have suggested that the right is protected under the 9th Amendment which states that even though the constitution enumerates some rights, this should “not be construed to deny or disparage others retained by the people.” However, the substantive DP doctrine has received the most traction for affording American’s a constitutional right to privacy. This theory claims that the right to privacy is also part of the seemingly ever-expanding list of rights protected under the “liberty” interest of the DP clause.
Griswold v. Connecticut, decided in 1965, was perhaps the first SC decision that recognized the right to privacy. The SC invalidated a CT law prohibiting the use of “any drug, medicinal article or instrument for the purpose of preventing contraception” arguing that a fundamental right to privacy is protected by the constitution that includes the right to use contraception. The majority used the “penumbra” from the Bill of Rights theory, while two concurring opinions used the 9th Amendment and substantive DP theories respectively.

In Eisenstadt v. Baird, the Court then extended the privacy right creating in Giswold, holding that single people have the right to possess contraceptives on the same basis as married couples. It is unclear how the Court determined that this was a fundamental right, given that at common law the view was that sexual partners had no legally enforceable rights without a marriage contract. Therefore, this was not a right that was “deeply rooted in the nation’s history.”

In Roe v. Wade (1973), the SC abused its discretion even further reading a general right to abortion into the broadened interpretation of “liberty” under substantive due process. Not only does this stray even farther from the initial intent of the DP clause which did not intend to cover a general right to privacy, but, as Rehnquist noted in his dissent, an abortion is not even a “private” right in the ordinary sense of the word. Even worse, by breaking pregnancy into trimesters and outlining when it is permissible for states to make restrictions, the Court essentially participated in a form of judicial legislation rather than strict interpretation of the intent of the drafter’s of the 14th Amendment. The Court had the opportunity to correct this aberration by overturning Roe in Planned Parenthood v. Casey. Instead, it reaffirmed Roe and created another unworkable framework in its unprecedented undue burden test. What is most troubling was the Court’s justification for not overturning Roe. It claimed that “a decision to overrule should rest on some special reason over and above the belief that a prior case was wrongly decided.” How can this be correct? Democracy requires thoughtful deliberation that leads to a sharp crystallization of the issues and honest self-assessment. The only thing worse than activist judges inventing policy rather than interpreting law is when those judges are incapable of admitting when they’ve made a mistake.

Conclusion
Our founders would have been flabbergasted by the idea that the Due Process clause, originally intended to protect procedural rights like a fair trial, would be interpreted by federal courts to protect abortion as a fundamental right. The American people are perfectly capable of using the legislative and executive branches of our government to change our laws to reflect the changing values of society. Issues currently being protected under the false theory of substantive due process should be left to the legislative branch, the one most capable of reflecting the will of the people. The judicial branch should return to its original purpose, using a disciplined, textualist approach to interpreting the most plausible reading of the law.

Posted in Law and Technology | No Comments »

Equal Protection: Who Should be Protected?

May 8th, 2006

The Problem
The EP clause of the 14th Amend says that “no state shall make or enforce any law which shall…deny to any person…the equal protection of the laws.” This clause was enacted shortly after the Civil War in order to ensure that ex-slaves were treated equally as other (white) citizens. However, its broad language has been interpreted as generally restraining or placing limits on government’s use of classifications in other areas, such as gender, nationality, etc. While the EP clause does not expressly limit the federal government, the DP clause of the 5th Amend has been interpreted to place the same restrictions on the federal government as the EP clause places on the states. One of the difficulties in interpreting the EP clause is that while it places restrictions on government use of classification, all laws necessarily classify groups of individuals in order to create incentives and disincentives in society. Therefore, identifying whether a class is to be protected and the level of scrutiny the courts should apply when judging the necessity of a particular classification is crucial.

Who should be protected?
Identifying which classifications or groups should be protected under the EP clause, and with what level of judicial scrutiny, is a difficult question. In addition to providing context to this discussion with some background on the history of equal protection, I will argue that the EP clause should be used to protect those groups that have been unconstitutionally deprived of an appropriate level of democratic influence such that a judicial check is necessary to ensure they are treated equally under the law. The reasons why these groups are unable to protect themselves appropriately through the political process may differ.

Early Equal Protection Decisions
The most obvious group requiring protection under the EP clause are racial minorities. The 14th Amendment was enacted in response to the gross inequalities afforded blacks compared to whites prior to and immediately following the Civil War. Therefore, the supermajoritarian will of the people clearly intended that racial minorities be protected. However, the question of what level of judicial scrutiny is appropriate still remains.

Early SC decisions began to define what equal protection actually meant. For example, in one of the first cases dealing with equal protection, Strauder v. West (1880), the SC held that exclusion of blacks from juries for no reason other than race was a violation of EP because the purpose of the clause was to assure that blacks had the same rights as whites, and that these rights should be protected by the courts. However, the SC did not always interpret the clause in a way that favored integration of blacks into the general community. In Pace v. Alabama (1883), the SC upheld a statute making interracial adultery a greater crime than regular adultery saying it did not violate equal protection because both races were punished the same. Moreover, in the Civil Rights Cases (1883), the SC created the state action doctrine limiting the EP clause to actions sanctioned by the state, leaving the private sector free to discriminate based on race. And in Plessy v. Ferguson (1896), the SC upheld the separate but equal doctrine, holding that arbitrary separation based on race did not constitute a badge of servitude.

Therefore, despite the intention of the 14th Amendment EP clause that similarly situated people of all races be treated equally under the law, blacks continued to be discriminated against (particularly in the South) in ways that did not reflect the will of the supermajority post-Civil War. Given this backdrop, the SC noted in a footnote in its US v. Carolene Products opinion, that a more stringent standard of review was necessary to protect racial minorities. Additionally, it seemed to implicitly recognize that the broad language of the EP clause did not seem to limit the protections afforded under the amendment to classifications based on race. Therefore, the SC further noted that “religious or national” minorities may be classes that require stringent judicial review under equal protection. However, it broadened the scope of the EP clause’s potential application even further by stating that any “discrete and insular” minority may need heightened review if they tended to be so disfavored that the political system would not work to protect them. In these cases, the opinion posited, the courts would have to make extra efforts to protect them.

Emergence of “Strict Scrutiny”
This was the foundation from which the “heightened scrutiny” standard emerged in Korematsu (1944), requiring that legislative classifications based on race be narrowly tailored to a government interest. This was expanded in Loving v. Virginia, where the Court declared that it would apply “rigid scrutiny” to racial classifications, and that they would not be permissible unless they were “necessary” to achieve a “compelling government objective.” There are still questions about the adequacy of these standards, given that racially neutral laws are typically upheld even if they have widely disparate impacts on racial minorities, Arlington Heights and Washington v. Davis. Some argue that because some racism is subconscious, and it is fairly simple to invent some non-discriminatory purpose (e.g. economic justifications, etc.) for race-neutral legislation that has discriminatory impact, Courts should have even more leeway to prevent equal outcomes to remedy these situations. Nevertheless, in general the Court’s recognition of heightened scrutiny has done a better job providing blacks equal protection than previous doctrines like separate but equal.

Gender Discrimination
In addition to racial minorities, women are another group that has historically been unconstitutionally deprived of an appropriate level of democratic influence such that a judicial check is necessary to ensure they are treated equally under the law. The prescient footnote to Carolene Products paved the way for this group to obtain heightened scrutiny. In the early 1970s, the SC began to acknowledge this need invalidating its first state law discriminating on the basis of gender in Reed v. Reed (1971) followed closely by Frontiero v. Richardson (1973). Finally, in Craig v. Boren, the SC clarified the standard of review for gender classifications into what is currently known as Intermediate Scrutiny. Under this level of scrutiny, the Court held that a gender-based classification must have a substantial relation to achieving an important governmental objective.

Is Protecting Discrete and Insular Minorities Enough?
The Court was correct in interpreting the EP clause as providing varying levels of judicial scrutiny depending on what is needed to ensure a particular group is treated equally under the law. It rightly recognizes that the need to scrutinize these classifications must be balanced against valid state interests that may require these classifications. However, the Carolene Products claim that discrete and insular minorities are the only groups that may need this additional level of judicial scrutiny is both under-inclusive.

First, it is under-inclusive because it fails to recognize that politically ineffective majorities can also exist. For example, as discussed in his Law Review article, Bruce Ackerman provides a hypothetical example of a black majority that is underrepresented in the political system by a dominating white minority. My initial argument that any group that has been unconstitutionally deprived of an appropriate level of democratic influence should receive a higher level of judicial scrutiny would cover this situation as well.

Second, it is under-inclusive because it fails to recognize that in many cases, judges should protect groups that are the opposite of discrete and insular. Groups that are discrete and insular gain political advantage by being close together enabling them to form well-organized lobby’s to press their concerns in the political process. The transaction cost for groups that are diffuse to form comparable lobbying groups are much higher. Moreover, in a tightly held community, a minorities non-participation in the political lobbying process will be more apparent and there will be more pressure to participate. On the other hand, in diffuse and anonymous groups there are larger problems of free-riding. They will not have the pressure to participate that comes from being in a single community, and their anonymity (the fact that they can hide their classification from others) means that they have less incentive to push their political agendas. Homosexuals are a perfect example of a diffuse group whose individuals can, if desired, remain anonymous. As a result, they are less likely to form influential lobbying groups to ensure they are adequately protected by the political process than discrete and insular groups like racial minorities. Therefore, Carolene Products declaration that discrete and insular groups, in particular, require a higher level of judicial scrutiny is not enough.

Conclusion

The EP clause should be used to protect those groups that have been unconstitutionally deprived of an appropriate level of democratic influence such that a judicial check is necessary to ensure they are treated equally under the law. While providing protections for discrete and insular groups like racial minorities, and discrete/diffuse groups like gender is a step in the right direction, it is not enough. The equal protection clause must also protect diffuse and anonymous groups such as homosexuality with a higher level of judicial scrutiny as well.

Posted in Law and Technology | No Comments »

Affirmative Action - Comments on Existing Caselaw

May 8th, 2006

What is Affirmative Action?
Affirmative action is when government programs are created to try to remedy past discrimination or promote general diversity by providing minority groups special considerations compared with the rest of society. These special considerations may include access to jobs, promotions or admission to universities.

The Problem?
Some argue that affirmative action is merely reverse discrimination sanctioned by the government that effectively excludes more qualified individuals based on their race and violates the majorities EP under the law. Others argue that the country has a compelling interest in remedying past discrimination and promoting a diverse culture that improves the quality of society. They would further argue that since the EP clause was meant to protect minorities from being treated worse than the majority, non-invidious racial classifications aimed at helping minorities, including affirmative action, should not be prohibited.

The issue, then, is finding the appropriate balance between these competing opinions as a guide to determine when affirmative action is appropriate.

The Correct Balance
Wherever this balance lies, the starting point is to ensure that when affirmative action is used, it is actually effective in achieving its dual aim of remedying past discrimination and promoting general diversity. Its application must not be over-inclusive such that it helps members of a minority group that may not need the help, such as rich black families at the expense of white families that may be more in need. Moreover, it must not reduce the incentives of the minority to perform optimally by encouraging qualified minorities to slack off, knowing they are likely to still be afforded opportunities based on their race. Conversely, affirmative action must not be used in ways where it reduces the incentives of the outperforming majority to perform optimally by feeling that their hard work will not be as fruitful given the special considerations given to the minority. Additionally, affirmative action must be applied in a way that is not condescending and demeaning to minorities by effectively saying they are incapable of earning the opportunities on their own merits. And similarly, it must not be applied in a way that breeds resentment between the majority that may be outperforming without the special treatment.

Given these concerns, affirmative action must be strictly scrutinized to ensure that the means used are narrowly tailored towards achieving the specific goals of remedying past discrimination and promoting diversity. With a few exceptions, the SC has balanced these concerns admirably.

Affirmative Action in Business
First, the SC has recognized the importance of strict scrutiny in reviewing race-based classifications used for the purposes of affirmative action. In City of Richmond v. Croson, the SC held that any affirmative action program that classifies on the basis of race will be strictly scrutinized, regardless of which race is burdened or benefited by the classification. The court held that a city cannot adopt a set-aside program that favors minority owned contractors where there was no evidence of discrimination because the 14th Amend. EP clause limits states’ ability to use race as a criterion for legislation. While Richmond showed that minority business received less than 1% of prime contracts despite representing about 50% of the general population, the SC said that statistical generalizations can’t substitute for evidence of discrimination. Moreover, when creating plans narrowly tailored to remedy past discrimination, the SC said the government cannot use broad statistical requirements simply to minimize associated administrative burdens of managing these plans.

This rule that all racial classifications must be narrowly tailored to further compelling state interests was extended to the Federal Government in Adarand Constr. v. Pena. In Pena, the US DOT awarded a bid to a subcontractor that was certified as a small business controlled by socially and economically disadvantaged individuals. A different subcontractor, that submitted the lowest bid, sued the federal gov’t. The SC held that strict scrutiny applies to all race-based actions and that the gov’t can only attempt to redress past discrimination in ways that are narrowly tailored.

While critics may argue that Congress is different from state legislatures because it is less likely to be influenced by local prejudices, these two decisions represent good policy. In a business context, affirmative action is not typically about promoting general diversity for the benefit of society. It is about remedying past discrimination. However, the Court accurately recognizes that providing special considerations to minority-owned businesses is not narrowly tailored to achieve this goal. Therefore, the Court was correct in holding that set-aside programs for minorities are unconstitutional when there is no evidence of discrimination. Moreover, it is not narrowly tailored because just because a business is owned by a member of a minority class does not mean that the proceeds and benefits of the company are going to benefit the minority group. A company made up of entirely white employees could simply hire a minority owner as a figurehead to be able to take advantage of the special opportunities afforded to this special classification. Additionally, this strict scrutiny ensures that businesses are properly incented to perform optimally and does not breed resentment that underperforming companies will be afforded better opportunities based on the race of its owners.

Therefore, the SC was right to strike down these statutes because the purpose of affirmative action in a business setting is primarily to redress past discrimination, and affording minority-owned businesses special advantages is not narrowly tailored to this goal.

Affirmative Action in Schools
Affirmative action in a school setting has two purposes. First, it is used to remedy past discrimination in which minority groups were not afforded the same access to education as whites. Second, it is used to promote a diverse culture in the education system to benefit society as a whole.

In Grutter v. Bollinger, decided in 2003, the SC upheld the University of Michigan Law School’s affirmative action admissions policy. In a 5-4 decision, the Court upheld its precedent in Regents of the University of California v. Bakke, stating that while racial quotas are unconstitutional, educational institutions can legally use race as one of many factors in their admissions process. A quota, or other mechanical formula was not considered to be narrowly tailored to the compelling interest of maintaining a diverse educational system.
In Gratz v. Bollinger, heard at the same time as Grutter, the SC struck down an undergraduate admissions policy based on points saying it was too mechanistic and therefore unconstitutional. In this case, an undergraduate needed 100 points to be guaranteed admission. There was a 20-point bonus for blacks, Hispanics and native Americans. The SC basically deemed the policy a quota system.

Critics of Grutter and Gratz say that while the SC claims to use a strict scrutiny analysis, it actually is not because the desire for racial diversity is not a compelling state interest. However, the EP clause is supposed to be a judicial check that ensures groups that have been unconstitutionally deprived of an appropriate level of democratic influence are treated equally under the law. Therefore, it is for the Justices to decide, as part of this judicial check, what is a compelling state interest.

Moreover, the Supreme Court reached a good balance in these decisions. Prohibiting quotas and other mechanical formulas is good because those systems often suffer from the over-inclusive problem. They tend to benefit the members of the minority class that are wealthy and educated and are in less need of help. Moreover, prohibiting quotas reduces the risk of breeding resentment amidst well-performing individuals in the majority because it is harder to find clear examples of reverse discrimination. Furthermore, by upholding using racial classification as one of many factors in the admissions process, it provides the schools an avenue for pursuing the goal of creating a diverse educational experience for its students. It also provides some leeway to redress specific instances of discrimination where they are warranted.

The Supreme Court Goes Too Far
In Johnson v. California, decided in 2005, the SC stated that racial classifications receive close scrutiny even when the classifications are meant to benefit the races equally. Here, a prison used race to determine which prisoners to pair up in cells due to major problems with racially-based, gang-related violence. The SC held that the standard of review should be strict scrutiny and that this classification was immediately suspect, even if it was intended to benefit both of the races. Although the SC remanded the case back to the lower courts, it is likely that their decision will force California to alter its practice of segregating by race.
In situations where the racial-classification benefits both classes, a lower level of judicial scrutiny is warranted. The state’s purpose in preventing prison violence is certainly compelling. Moreover, the goal of promoting diversity in a prison setting is significantly less compelling than in the educational system.

Conclusion
When race-based classifications used in affirmative action benefit one race at the expense of another, they should be strictly scrutinized and narrowly tailored towards the compelling goals of promoting diversity in society and redressing past discrimination. When race-based classifications benefit both races equally, the SC should use a lower level of judicial scrutiny.

Posted in Law and Technology | No Comments »

Google Book Search - Potentially Liable for Direct Copyright Infringement?

March 14th, 2006

Note: access to an entire paper written on this subject is available here: GOOGLE BOOK SEARCH, BITTORRENT AND COPYRIGHT LAW

An Assessment of Google’s Potential Liability for Direct Copyright Infringement

i. Direct Copyright Infringement

As discussed in the Technical Background, Google Book Search has two key facets: the Print Publisher Program and the Print Library Program. The Print Publisher Program does not present any copyright issues because under this program Google obtains authorization from copyright holders to scan the text of copyrighted books into Google’s search engine. However, under the Print Library Project, Google has partnered with major libraries to scan books without the express permission of the original publishing companies. The scanning of copyrighted books and creation of multiple digital copies (one for the libraries and one for Google’s search index) constitutes direct copyright infringement.

There are three aspects of the project that could present copyright issues: 1) the digital copy Google scans for use in its search database, 2) the display of a portion of that scanned text to users in search results, and 3) the creation of a digital copy of the books for the libraries. It is pretty clear that creating a digital copy of an unauthorized book for use in its search database is a copyright violation, unless a fair use exception applies. Similarly, the display of a portion of that scanned text to users in search results constitutes a fair use exception unless the fair use exception applies. The creation of a digital copy for use at the libraries is not as clear cut. If the libraries, in purchasing the books for their library, purchased the rights to create digital copies for archival purposes then contracting out to have an organization like Google create these copies for them does not seem to be a copyright violation. However, if the libraries do not have the right to make a digital copy from their hard copies, then this too constitutes a copyright violation. However, it is more likely that this would be seen as a violation on the part of the libraries rather than Google. Therefore, for the fair use analysis, this comment will look only at the first two examples of direct copyright infringement. Therefore, the primary legal question is whether Google’s Print Library Project, and in particular the creation of unauthorized digital copies of copyrighted books for use in its search engine, constitutes a fair use of the copyrighted work and therefore escapes liability under the fair use exception outlined in Section 107 of the United States Copyright Act.

ii. Fair Use Defense

In order for Google’s Print Library Project to escape liability under the fair use doctrine, an analysis of the elements of the fair use test must be made. In particular, when deciding whether a fair use exception exists, the courts must take into account 1) the purpose and character of the use, including whether the use is of a commercial nature or is for nonprofit educational purposes, 2) the nature of the copyrighted work, 3) the amount and substantiality of the portion used in relation to the copyrighted work as a whole, and 4) the effect of the use upon the potential market for or value of the copyrighted work.
Regarding the purpose and character of Google’s use, the court in Arriba asked whether the use was commercial in nature. It found that since Arriba’s use of Kelly’s images was not “highly exploitive”, the commercial nature of the search engine only weighed slightly in favor of Kelly. Regarding the display of sentences of surrounding text to end users in search results, Google is attempting to profit from providing these portions of copyrighted material. First, Google will eventually use advertising on these pages. Therefore, similar to Grokster in which the Supreme Court found that the P2P file-sharing companies had found a way to convert users copyright infringement into dollars, Google has found a way to profit off of the unauthorized digital reproduction of copyrighted work.

Nevertheless, Google’s Book Search is a tool that will bring the existence of relevant books to the attention of potential users. Therefore, despite the purpose of its infringement being commercial in nature Google has a fairly strong public policy argument suggesting that its purpose is for the general public good and will not act to supplant the revenue streams of the copyright holders themselves. Rather than being of the character or purpose that would directly compete with sales of the books themselves, Google Book Search acts to increase exposure to books and increase revenue for publishers. This is similar to Arriba in which the court held that Arriba’s use of Kelly’s images promote the goals of the Copyright Act, and the fair use exception in particular, by benefiting the public by making information more readily available. Moreover, it held that Arriba’s use of thumbnail versions of Kelly’s images did not supplant the need for the originals because there was not enough value that could be gleaned by the low quality thumbnails themselves. Similarly, most users of Google Book Search will not use this tool as an alternative to buying the book itself. The value they receive from getting a few sentences of material, in most cases, will not be enough to supplant book sales in the same way that low quality thumbnail pictures will not supplant the need for the originals. Therefore, despite Google benefiting by a larger stream of advertising revenue through its use of these copyrighted works, this would be counterbalanced by the public policy benefits of having this content readily available to the general public and the fact that the model would be more likely to drive revenue for publishers rather than supplant their existing revenue streams. Therefore, it is unlikely that the purpose and character of Google’s use, though commercial in nature, would weigh heavily against Google.

Second, the nature of the copyrighted work must be considered. Here, the court in Arriba said that one of the core aspects of this element of the fair use doctrine was to ensure the protection of works that are creative in nature rather than “more fact-based works.” According to Jonathon Band, the “vast majority” of the works scanned under the Print Library Project will be non-fiction “fact-based works” rather than creative fictional works. Moreover, according to the court in Arriba, published works are more protected by the fair use doctrine because the “artist’s expression has already occurred.” Therefore, given that the vast majority of the Print Library Project will be to scan published non-fiction books that are more fact-based than fiction, this element will only weigh slightly against Google.

Third, the amount of the copyrighted work as a whole that is used must be considered. In many cases, quotes are used from copyrighted text. This constitutes fair use partially because the portion of the work that is used is not a substantial part of the whole. In the case of Google, the two separate copyright considerations require slightly different analysis. The full reproduction of copyrighted text and the storing of entire digital copies of these texts in their search databases seems to weigh heavily against Google. However, the distribution of a few sentences of that text that match users search criteria seems to be of the same nature as quoting copyrighted work that is generally accepted as fair use. In Arriba, a full copy of the copyrighted work is made available, albeit in lower quality. The court held that “if the secondary user only copies as much as is necessary for his or her intended use” that this factor of the fair use doctrine would not weigh heavily against them. It could be argued that in order to fulfill the purpose of providing instant, searchable access to all of the world’s books, Google can only achieve this with a full, digital copy of the text. Searches restricted to portions of text would not be as reliable, accurate and relevant as full text-based searches. Nevertheless, Google will have difficulty arguing that the full reproduction of copyrighted text in their search database is not a “substantial portion” of the copyrighted work “as a whole.” Therefore, despite the public policy benefits, this factor is also likely to weigh against Google.

Regarding the fourth factor, there are two interesting arguments. Google can argue that, if anything, its program increases the potential markets for the copyrighted works. The availability of full text-based search could increase demand for some books which could cause some users to buy books they otherwise would not have purchased. While the display of a few sentences of text around the users search criteria probably would not be enough to reduce demand for purchasing the entire book, this becomes a slippery slope. Google has incentive to provide as much of the copyrighted work as would be authorized under the law. The publishers would be incented to allow Google to provide just the right amount of information to pique the user’s curiosity, make clear that the content is relevant to the users needs, and encourages the purchase of hard copies. However, even if providing a few sentences of content through Google Book Search would not significantly displace sales of new books, it does impede a publisher’s ability to license new digital copies of its content to libraries or to search engine providers like Google. For example, in scanning books from libraries Google is essentially paying for the rights to fully reproduce copyrighted works in its search technology with a digital copy the libraries can use for its archival purposes. Therefore, Google’s method of payment is, in effect, an illegal reproduction of a copyrighted work.

Moreover, Google has an adequate alternative. Google could pay publishers for the right to provide this search feature on their website. Moreover, if Google were not creating these types of contracts with libraries, these publishers would have the ability to sell digital copies of their content to libraries for archival purposes themselves. Therefore, by pursuing the unauthorized reproduction of copyrighted work and giving a copy of those works to the libraries providing access to books, Google cuts off two possible modes of revenue to the copyright holders themselves. Therefore, even if Google’s technology drives new revenue by making books more accessible online, it is not clear that this increase in revenue will outweigh the loss in potential revenue to publishers who would otherwise be able to charge libraries for digital copies and license the content to search providers for inclusion into tools like Google Book Search.

If copyright is to mean anything, companies like Google cannot be authorized to make digital copies of entire works that they have not purchased for commercial gain. Therefore, given that all four of the fair use factors seem to weigh against Google, it is unlikely that Google Book Search will be protected from liability under copyright law under a “fair use” exception. Rather than circumventing the publishers by going directly to libraries for access to copyrighted materials, Google should negotiate with publishers to license access to their works for inclusion in its Google Book Search feature.

Posted in Law and Technology | No Comments »

BitTorrent - Could it Have Been Liable for Copyright Infringement?

March 14th, 2006

Note: access to an entire paper written on this subject is available here: GOOGLE BOOK SEARCH, BITTORRENT AND COPYRIGHT LAW

An Assessment of BitTorrent’s Potential Liability for Copyright Infringement

1. BitTorrent’s Potential Liability Under an Inducement Theory

In Grokster, the Court laid out the test under the inducement theory stating that a company is liable for contributory infringement if they 1) intend to bring about infringement, 2) by distributing a device suitable for infringing use, and 3) actual infringement occurred. Given what some have called the “Miss Manners” rule, it seems that to avoid liability under an inducement theory companies must simply avoid marketing their products as ones capable of direct copyright infringement. Companies will begin setting up processes to ensure their internal communications and externally facing documentation could not be interpreted as having wrongful intent in distributing their product. Therefore, this rule makes it increasingly difficult to tell the difference between companies that truly don’t want infringement to take place and those that do want it, but are able to sufficiently hide their intentions. So how will BitTorrent fair under this new rule of secondary liability? BitTorrent should be protected. As discussed in the Technical Background, BitTorrent is a P2P technology that is very good at sharing large files. One study claims that in 2005 over 35% of all Internet traffic was taking place over the BitTorrent network.
BitTorrent is different from Napster, Aimster and Grokster partially because BitTorrent refers to both 1) the communication protocol and 2) a default application employing the protocol to share files. It is clear that the the BitTorrent company should not be liable, under any theory, for the mere creation of a new communication protocol. However, the question is whether they would be liable under an inducement theory for the distribution of a P2P file-sharing application that employs that protocol. While there is no doubt that BitTorrent is an application that is suitable for infringing copyrights and that actual infringement on the part of its users occurs, a court should not find that they have the requisite intent needed to establish secondary liability under the inducement theory.

BitTorrent’s original customer was etree, a community of pepole who distribute live concert recordings for free online, with the permission of the bands. Bram Cohen, BitTorrent’s founder has said that he developed BitTorrent based on the needs of his friends to legally share and download these types of shows. Additionally, there have been other significant commercial uses of the technology. The protocol has been used to distribute patches for Blizzard Entertainment’s World of Warcraft game and many software companies use it to distribute builds among their technology teams. In Grokster, the companies distributing file-sharing applications were found liable under an inducement theory because they 1) specifically sought to capture Napster’s market of known copyright infringers, 2) they actively induced their users to infringe through their instruction manuals and customer support networks, 3) failed to filter out copyrighted materials, and 4) profited from the infringement through advertising revenue. First, BitTorrent has not tried to capture a pre-existing group of copyright infringers from a different technology. Rather, from the very beginning Bram Cohen has made it clear that his intent is to serve a legitimate market of providing tools for legally sharing and downloading concerts. Second, BitTorrent does not seem to induce infringement through their documentation. Their FAQ section specifically outlines their goals and purpose. Even a very early version of the BitTorrent Frequently Asked Question (FAQ) list said that BitTorrent was developed for legal trading of these types of shows. Moreover, when the company was testing the initial versions of the technology, they used freely available, non-copyrighted video content. By contrast, the examples in Grokster’s documentation all involved infringing content.

Bram Cohen did write, in his Technological Activist’s Agenda, that he “build[s] systems to disseminate information, commit digital piracy”, etc. However, while this statement was available from the BitTorrent website for about two years from July 2001-July 2003, a lawsuit under active inducement should not read much into this statement. Sony also created a system that could be used to disseminate information and commit digital piracy. Simply saying that the system can be used to commit digital piracy is not the same thing as actively inducing others to use the product in this way. Third, while BitTorrent does not provide tools for filtering out copyrighted material, this should not weigh heavily against BitTorrent. Courts and judges are not in the best position to judge the quality and efficiency of a particular product design. They are not well-positioned to be making cost-benefit analyses of different feature sets or product designs. Therefore, while the Grokster court did mention this as evidence of Grokster’s intent, the court’s are likely to limit the impact of these types of product design questions on their decisions of finding intent. Fourth, while Grokster and Streamcast were profiting from infringement through translating demand for copyrighted materials into advertising dollars, BitTorrent is entirely free and does not include advertising in its default client application. That being said, BitTorrent recently added a search engine to their website that does include paid advertisements. This will probably be the biggest hurdle for BitTorrent to overcome in proving they were not inducing or encouraging infringement. In the same way the Court found that Grokster was converting demand for copyright infringement into advertising dollars, a court could similarly find that BitTorrent’s new advertising based search capabilities serve the same purpose. Moreover, BitTorrent’s recent move to a decentralized, trackerless technology design may make it more difficult to escape liability. While the court never specifically faulted Grokster for developing a decentralized technology in an attempt to escape liability, a court could find that movement in this direction was evidence of intent to have its users infringe. In the same way Grokster tried to avoid liability with its decentralized use of Supernodes to avoid centralized index files there is an argument that BitTorrent’s move to trackerless torrents could be seen as evidence of intent to have its users infringe.
However, the overwhelming evidence is that BitTorrent has not “shown by clear expression or other affirmative steps” that they intend for their users to infringe copyright. The consistency by which BitTorrent’s owner, Bram Cohen, has stated BitTorrent’s legal purpose and target markets and the lack of documentation or tools for helping users use the technology to commit infringement should be enough to protect BitTorrent from liability under an inducement theory. Nevertheless, the new inducement theory creates less predictability for developers and technology companies. It increases the cost of the entire legal process by making discovery more expensive as it requires more investigation to discover and identify a product company’s true intent.

2. BitTorrent’s Potential Liability Under Contributory Copyright Infringement

While both Napster and Aimster were found liable for contributory copyright infringement, the circuit courts had two different tests that emerged to determine whether the defendant had the requisite knowledge required by the second element of the “staple article of commerce” doctrine adopted from patent law in Sony. In the Ninth Circuit, drawing heavily on Napster’s centralized architecture the bright line test required the prosecution to show that 1) the defendant had reasonable knowledge of specific infringing conduct, and 2) the defendant failed to act to prevent the distribution of infringing works. In the Seventh Circuit, the court would apply Posner’s balancing test that would weigh the evidence including 1) percentage of noninfringing to infringing use, 2) whether the defendant “invited” infringement (e.g. in its documentation and tutorials), and 3) whether defendant took steps to eliminate infringing uses. While the Supreme Court did not address this issue directly in Grokster, Ginsberg and Breyer’s two concurring opinions discussed very different views on whether a technology company could be liable based solely on the distribution of a product capable of infringing use. Breyer’s rule is better at promoting technological innovation and protecting the interests of future, undeveloped markets that could benefit from emerging technologies.

i. Liability under Ginsberg’s Balancing Test

Under Ginsburg, BitTorrent would have a harder time escaping liability. Ginsberg more closely followed the Seventh Circuit balancing test saying that a technology company can be liable for contributory infringement based on distributing a product capable of infringement alone. A company that distributes a product that is not capable of substantial, commercially significant, non-infringing uses would be liable for the copyright infringement of their users. Therefore, BitTorrent would have to show that its technology had substantial, non-infringing use that were not simply the “anecdotal” evidence she saw in Grokster. She recommends the balancing test of looking at the percentage of infringing and non-infringing uses in order to determine whether they are “substantial.” Therefore, this test would essentially make the test for secondary liability based on contributory copyright infringement a predominant use test. If the software is predominantly used for infringing uses, it is liable for contributory infringement. If the software is predominanly used for non-infringing uses, it escapes liability. BitTorrent would probably fail this test because currently, a large portion of its users use the technology to download copyrighted movies and television shows. It would be more difficult for Ginsberg to simply dismiss the evidence of non-infringing BitTorrent uses as being “anecdotal.” For instance, unlike Grokster, BitTorrent was created with a lawful, non-infringing use in mind - sharing and downloading concerts with the permission of the bands. Moreover, there are a number of examples of commercial uses of BitTorrent to distribute patches for video games and software, as well as a tool for distributing large applications like Linux or other operating systems. Nevertheless, when comparing strict percentages the infringing use very likely dwarfs the non-infringing use.
This is troubling given the vast potential for the technology. Some people consider BitTorrent to be the world’s largest TiVo. They believe it is a technology that will revolutionalize the distribution and production of content as more content providers model their creative works for online audiences. One example was a graduate computer science student who enjoyed Outfoxed, a documentary that criticized Fox News. The student convinced the film’s producer to allow him to host the file online as a .torrent file. 1,500 people downloaded it in the first two months, effectively turning this computer science student into what he called “a movie producer.” “If I had my own content, I’d be a TV station.” A similar example was when Jon Stewart, host of The Daily Show, appeared on CNN’s Crossfire and verbally insulted the hosts calling them political puppets. The content was made available as a .torrent and within a single day 4.000 servers were swapping the clip. Over two million people streamed it from a website over the next few weeks while only ~850,000 people saw the live broadcast. More and more people will begin watching TV and film content online. However, for these changes to take place the technology facilitating these revolutions must be protected.

If the correct interpretation of Ginsberg’s balancing test is not simply a predominant use test, but rather requires the more careful balancing of potential future markets, development plans to reduce infringement, etc., the test is still inefficient. Under this understanding of Ginsberg’s approach balancing test, BitTorrent would have to provide detailed evidence of its business plans, future market opportunities, projected feature development cycles and potential revenue streams in order to provide the court with the information needed to perform the balancing test. It must show convincingly that the evidence of its commercial uses are more than the examples Ginsberg considered “anecdotal” in Grokster. While this will provide BitTorrent more leeway in proving its potential value than under a strict predominant use test, the Court would still be likely to find BitTorrent liable for contributory infringement.

ii. Under Breyer’s “Capable of” Substantial Non-Infringing Uses Approach

BitTorrent would almost certainly be excused under Justice Breyer’s approach. Under Breyer’s understanding of Sony, granting Grokster summary judgment based on a theory of contributory infringement for distribution of a product capable of infringing use was appropriate. He argued that Sony passed this test because Grokster was “capable of” substantial, non-infringing uses and that this was not a balancing test as Ginsberg would like. He argued that the ability to use Grokster to share authorized music, free electronic books, and licensed video files were all evidence of non-infringing uses that could reasonably be expected to expand over time. This same argument holds true for BitTorrent. Therefore, under Breyer’s rule BitTorrent would be free from liability under contributory infringement because as more and more non-copyrighted or licensed material became available in digital form, it would be reasonable to assume that lawful P2P file-sharing would become more prevalent.

This is a better rule than Ginsberg’s balancing test. Breyer’s reading of Sony does a better job of protecting the interests of users and creators of P2P software while avoiding the problems associated with implementing Ginsberg’s balancing test. Ginsberg’s balancing test would lead to inconsistent rulings as different judges weighed the benefits and costs differently based on their subjective valuations. Moreover, Breyer’s understanding of the Sony rule better protects technologies that may have undeveloped future markets. The nature of technology is that its innovation forces companies and individuals to reconsider the best, most efficient means of executing on their business plans. Technologists should be encouraged to invent without the fear of being dragged into court for the mere creation and distribution of a tool that might initially be used primarily to infringe copyright. Second, this rule is more effective because it recognizes the limitations judges have in answering technical questions about the commercial viability of different products or product features. Breyer’s rule recognizes the important distinction that technology does not commit infringement, people do. Rules of contributory infringement must protect the interests of technologists creating and innovating in order to find new solutions to problems. The burden should be on copyright holders to develop appropriate means of protecting their rights. As Breyer’s points out, if the movie and television industries are frustrated at the amount of their content being distributed over BitTorrent they are perfectly capable of creating new technologies like “digital watermarking” or “digital fingerprinting” that could be used to prevent the misuse of their content.

Moreover, if they did develop these types of technologies, Breyer’s rule would protect them against tools that would decode their encryptions in order to remove their digital rights management technologies. Not only is this type of obstruction covered in the Digital Rights Management Act, but it is similar to descramblers and other technologies that courts have held do not have “substantial” non-infringing uses.

Posted in Law and Technology | No Comments »

Google Book Search - Legal Background

November 30th, 2005

Note: access to an entire paper written on this subject is available here: GOOGLE BOOK SEARCH, BITTORRENT AND COPYRIGHT LAW

Here is background on one of the leading cases addressing copyright issues in the context of search engines.

1. Introduction to Legal History
After search engines began to be used prevalently to organize and help find the world’s digital information, legal disputes soon followed. Many lawsuits had to do with trademark issues. For example, search engines were dragged to court by company’s that were frustrated that the search engine sold advertising rights to one of their competitors based on search criteria that included that company’s trademarks. Moreover, search engine caching and the indexing of copyrighted web content have led to the question of what constitutes fair reuse of copyrighted web content.

2. Kelly v. Arriba Soft Corp. 336 F.3d 811
After Napster, the Ninth Circuit was forced to address a different type of fair use. In this case, Kelly was a photographer who made his copyrighted pictures available over the Internet. Arriba Soft took Kelly’s pictures and made thumbnails, small low quality versions of digital pictures, which they made available through their search engine. Arriba’s database did not store the full picture but only the thumbnail. Clicking on the thumbnail would take the user directly to Kelly’s website, in an Arriba Soft web frame, where the full picture was made available. Kelly sued Arriba for direct copyright infringement. The lower court found that Arriba Soft violated the copyrights and did not have a valid fair use defense for providing the thumbnails from within their search engine.

On appeal, the Ninth Circuit reversed holding that the thumbnails did constitute fair use. Therefore, Arriba Soft was allowed to display copies of images (acceptable resolution was not decided) in their search results. In analyzing the fair use doctrine, the Ninth Circuit used Section 107 of the United States Copyright Act as its guide. Regarding “purpose and character of the use,” the court held that the use was commercial, but not of the same character as the original work. The images were not being sold and weighed in Arriba’s favor because there was “public benefit” to the search engine. Regarding the “nature of the copyrighted work,” the court held that the nature of pictures as published creative work was a factor that worked slightly in favor of Kelly. Regarding the “amount and substantiality” of the portion of the work used, the court said that it was necessary for Arriba to copy the entire image in order to allow their users to determine whether they wanted to pursue the link and obtain more information. Therefore, this factor did not weigh heavily in either parties favor. Regarding the “effect of the use upon the potential market for or value of the copyrighted work” the Ninth Circuit found that the transformation of the images for use in the search engine would not have an adverse effect on Kelly’s market for his images. Rather, it found that the thumbnails would guide larger number of users to Kelly’s work rather than away from it. The thumbnails themselves would not detract from this market because the size and quality would make them unattractive as a substitute for the original images.

Concluding that two of the fair use factors weighed heavily in Arriba’s favor and only one weakly favored Kelly’s position, the Ninth Circuit carved out a fair use exception for using thumbnails of copyrighted images in search engine databases.

Posted in Law and Technology | No Comments »

Google Book Search - Background on the Idea

November 30th, 2005

Note: access to an entire paper written on this subject is available here: GOOGLE BOOK SEARCH, BITTORRENT AND COPYRIGHT LAW

1. Quick background on Search Engines
Here is a quick primer on the evolution of search engine technologies. A search engine works by storing content on a large number of web pages and delivering search results to users based on search criteria. Search engines achieve this goal by 1) Crawling, 2) Indexing, and 3) Searching. Search engines send out crawlers, or web spiders, which retrieve pages from the web and follow every link that it finds. The content on each page discovered is then analyzed and indexed based on its content. This content is stored in the search engine’s index database. Moreover, some search engines “cache” the pages they crawl storing all of the content, or some portion of the content, in the database as well. These cached pages can be made accessible directly from the search engines website without hitting the original website.

Users of the search engine type in queries that typically includes words or phrases. The search engine looks for these “search criteria” in its index and, using a variety of different styles of algorithms, returns the web pages or content that best match the criteria. Often these search results include other information on the returned content such as its location, a description, some subset of the content, etc. Search engines become more useful the more relevant the search results it returns are to the user. Part of this is a function of the matching algorithm and part of it is based on the amount of information that is indexed.
In 2001, the Google search engine became popular because its PageRank algorithm was very good at providing search results that were relevant to the user.

2. Google Book Search
OK, so what is this whole Google Book Search thing about? Google Book Search is an initiative by which Google plans to scan the full text of books into its search database. The initiative is broken down into two key areas: the Print Publisher Program and the Print Library Program. Under the Print Publisher Program, Google obtains authorization from copyright holders to scan the text of copyrighted books into Google’s search engine. When a user enters a query into the Google search engine, the engine scans its indexes and returns the user a set of bibliographic information on books that match the user’s search terms. Included with the search results would be a few pages of the book surrounding the text matching the user’s search criteria. Links are made available so that the user can purchase the book from the publisher or other online retailers. The Print Publisher Program does not appear to raise any copyright issues because under this program Google obtains a specific grant to the content directly from the copyright holder.

Under the Print Library Project, Google has partnered with libraries from major educational institutions like Harvard, Oxford and Standford to scan the text of the books within those libraries without the permission of the original publishing companies. In exchange for providing the libraries with a digital copy of the materials, Google then indexes a second digital copy for use within its search engines. When a user enters search criteria that match information contained in one of these books they provide results that include a few sentences of surrounding text. Therefore, for books that are still subject to copyright restrictions, Google restricts the information provided to users. Under this program, Google has also instituted an “opt out” policy in which the publisher can provide Google a list of books that it does not want them to scan from the libraries archives. Therefore, publishers are given the opportunity to prevent any of their copyrighted materials from being made available through Google’s search technology. The primary legal question is whether Google’s Print Library Project violates copyright or whether it is protected as a fair use of the copyrighted work.

Posted in Law and Technology | No Comments »

The Longtail of BPM

November 15th, 2005

The phrase “The Long Tail” was first coined by Chris Anderson in a 2004 Wired Magazine article. He observed that demand for niche products with low sales volume can, in aggregate, rival and sometimes exceed the demand for a few bestselling products or blockbusters. He noted that demand for these niche products make up the “Long Tail” of an overall product demand curve (see diagram). Anderson argued that traditional stores with limited shelf space must focus on selling mainstream hits or blockbusters because the high costs of inventory and distribution make serving niche goods in the long tail unprofitable. By focusing on products they know will have high demand, the companies maximize profit. However, Anderson’s primary thesis was that the Internet changed all of this. Companies like Amazon.com and Netflix, for instance, recognized that less popular goods (e.g. rare books or documentary videos) can also make money. These online retailers combine infinite shelf space with real-time information about buying trends and, as a result, can profitably serve these niche markets in addition to the blockbusters. Since there are so many more niche goods than blockbusters, the aggregate profit that can be squeezed out of each niche can quickly add up to impressive gains. For example, while the average Blockbuster store carries less than 3,000 DVDs, over a fifth of Netflix rentals are outside its top 3,000 titles. An Amazon employee described the long tail phenomenon by saying “We sold more books today that didn’t sell at all yesterday than we sold today of all the books that did sell yesterday.” And finally, Google figured out how to make money off the “long tail” of search terms, by providing an infrastructure for charging small prices for large numbers of niche advertising key words.

The long tail theory can also be applied to process automation and process management. Organizations have thousands of interconnected business processes. This patchwork of process logic includes human tasks as well as computerized activities that access and update enterprise systems and applications. While all of these processes work collectively to support the goals and strategies of the organization, enterprise software has typically been targeted toward high demand processes that are roughly similar across organizations (CRM, ERP, HCM, salesforce automation, etc.). For each process automated in one of these enterprise software applications there are hundreds of highly customized, unique organizational processes that are not adequately addressed by these systems. These processes are often being managed through e-mail, MS Excel, faxes and telephone calls or handled by custom coded software written by external consultants or internal IT organizations. Moreover, sometimes even the processes automated using rigid enterprise software packages are so unique or specific to the organization that they require heavy and expensive customizations that increase project risk and hinder upgrade paths. Given such costs, many organizations have not seen any significant value return on the implementation of these solutions. Some processes, such as strategic sourcing and procurement, are often so tailored to the specific needs of the particular organization that traditional off-the-shelf procurement packages will not suffice. Others, like human capital management systems and compliance solutions for Sarbanes-Oxley and HIPAA, require agile business logic that can change rapidly as regulatory statutes are reinterpreted by courts or modified by Congress. These processes could benefit from better management and automation if the right software were available at the right price, but historically the cost of developing custom software tomeet these niche, organization-specific, requirements has been too high given the low potential sales volume. Therefore, just as inventory and distribution costs makes selling niche products unprofitable in the world of brick-and-mortar stores, the cost of developing enterprise software solutions for organization-specific business processes is oftentimes cost prohibitive. This is all changing with the emergence of a new class of enterprise software: the Business Process Management Suite (BPMS). BPM Suites are rapidly becoming the preferred next generation platform for automating processes and building process-centric applications. By lowering the incremental cost of automating new business processes and modifying existing ones, BPM suites are helping organizations achieve efficiency gains from processes further down the long tail. Their enterprise performance and scalability greatly expand the number and complexity of processes that can be profitably automated within an organization. Once automated, these processes benefit from being well documented in a common language and notation and can be adjusted quickly to meet changing business requirements. Just as Amazon, Netflix, and Google have used the Internet to profitably serve niche markets, organizations can use BPM suites to drive efficiency gains further down into its core processes.

Posted in Business Process Management | 1 Comment »

P2P Filesharing - The Supreme Court Speaks in MGM v. Grokster

November 6th, 2005

Note: access to an entire paper written on this subject is available here: GOOGLE BOOK SEARCH, BITTORRENT AND COPYRIGHT LAW

Normative Summary
While both Napster and Aimster were found liable for contributory copyright infringement, the circuit courts had two different tests that emerged to determine whether the defendant had the requisite knowledge required by the second element of the “staple article of commerce” doctrine adopted from patent law in Sony. In the Ninth Circuit, drawing heavily on Napster’s centralized architecture the bright line test required the prosecution to show that 1) the defendant had reasonable knowledge of specific infringing conduct, and 2) the defendant failed to act to prevent the distribution of infringing works. In the Seventh Circuit, the court would apply Posner’s balancing test that would weigh the evidence including 1) percentage of noninfringing to infringing use, 2) whether the defendant “invited” infringement (e.g. in its documentation and tutorials), and 3) whether defendant took steps to eliminate infringing uses.

Procedural History
In Grokster, the music and motion picture industries brought an action seeking an injunction against two distributors of P2P software, Grokster and StreamCast (hereafter “Grokster”), under a theory of contributory infringement. This was the first time the Ninth Circuit was presented with a P2P copyright case involving technologies with decentralized architectures, different from Napster’s centralized index server approach. The District Court granted summary judgment in favor of Grokster saying it was not liable. In upholding summary judgment, the Ninth Circuit followed its analysis in Napster and applied its interpretation of the Sony doctrine to the knowledge requirement of contributory infringement. It therefore applied the tests from Sony and Napster and found that since Grokster was capable of substantial, commercially viable noninfringing uses, MGM had the burden of showing that the Grokster had reasonable direct knowledge of specific infringing conduct. The court held that, given the decentralized nature of Grokster’s product, the prosecution failed to show that Grokster had this direct knowledge. Therefore, the court held that Grokster was simply a distributor of software and that they did not materially contribute to the copyright infringement of their users. Having a decentralized technical architecture seemed to be the key for the software distributors to be able to escape liability. In addition to eliminating Grokster’s knowledge of direct infringement, it also prevented Grokster from having the ability to supervise or monitor how their end users were using the tool.

Supreme Court
Majority Opinion

When Grokster reached the Supreme Court, many believed the Court would use the case to decide the appropriate test for determining whether the knowledge element of the staple article of commerce doctrine requiring knowledge of direct infringement. The Ninth circuit’s more technology friendly approach stated that a distributor of software would only be held liable for contributory infringement if it had reasonable knowledge of specific infringement and that this knowledge does not exist with a decentralized P2P architecture. On the other hand, the Seventh Circuit’s more copyright holder friendly balancing test gave the court more discretion in weighing the pros and cons of the technology against the detriment of the copyright infringement. Which approach would the Supreme Court adopt?

Rather than answering this question, the majority opinion of the Supreme Court avoided the question altogether and decided the case by using another principle of patent law – the inducement theory. Under the inducement theory, the Court held that a software distributor that promotes the use of its tool to infringe copyright “as shown by clear expression or other affirmative steps taken to foster infringement” is liable for the resulting infringement of their users. Therefore, the Court held that the Ninth Circuit’s granting of summary judgment was inappropriate because they did not address whether Grokster took affirmative steps to induce its users to commit copyright infringement. The Court laid out the test under the inducement theory stating that a company is liable for contributory infringement if they 1) intend to bring about infringement, 2) by distributing a device suitable for infringing use, and 3) actual infringement occurred. The Court reconciled this new rule with the protection Sony provided under the staple article of commerce doctrine by saying that Sony dealt with the narrower issue of whether there is “a claim of liability based solely on distributing a product with alternative lawful and unlawful uses, with knowledge that some users would follow the unlawful course.” Rather, in Grokster the Court was addressing whether there is a claim when there is actual intent and affirmative steps taken to encourage copyright infringement.

The majority opinion rejected the Ninth Circuit’s interpretation of Sony that “whenever a product is capable of substantial lawful use, the producer can never be held contributorily liable for third parties’ infringing use…even when actual purpose to cause infringing use is shown…unless the distributors had specific knowledge of infringement at a time at which they contributed to the infringement, and failed to act upon it.” Rather, the Court held that the rule in Sony does not require courts to ignore evidence of the distributor’s intent. The majority opinion then applied this new inducement test to the facts in Grokster. Given that Grokster clearly distributed its product and that there was fairly substantial evidence that users used the software to commit actual infringement, the only element of the inducement rule the Court addressed at length was whether Grokster took affirmative steps to encourage its users to commit copyright infringement. The Court noted a number of areas that suggested that Grokster had induced infringement including 1) advertising and instructing how to infringe, 2) targeting former Napster users known to infringe, 3) failing to take steps to prevent infringement, 4) profiting from infringement, and 5) simply distributing a tool capable of infringement. Regarding advertising, the Court said that advertising an infringing use or instructing how to engage in an infringing use can show an affirmative intent that the product be used to infringe and can expose the software distributor to liability. The Court held that Grokster was not merely a passive recipient of information about infringing use. Rather, it voiced the objective that recipients use it to download copyrighted works, and took active steps to encourage infringement. The Court noted that Grokster distributed an electronic newsletter containing links to articles promoting its software’s ability to access popular copyrighted music.

The Court also discussed Grokster’s similarity to Napster, a tool used for infringement. Internal e-mails said they wanted to be positioned to capture the flood of Napster’s users that would be “actively looking for an alternative.” Furthermore, the CTO of Streamcast even said one of their goals was to “get in trouble with the law and get sued.” The Court even said the similarity of their product names was evidence of their intent to profit off of their users infringement. Regarding preventing infringement, the court noted that Grokster had made no attempt to use filtering technology to help discourage copyright infringement. According to the Court, this evidence underscored Grokster’s intentions to have their users commit copyright infringement. And finally, the Court noted that Grokster’s revenue stream was generated by advertising. Therefore, given that a large number of Grokster users (90%) were using the tool to commit copyright infringement, the Court found that Grokster was profiting based on its users infringement. It had “translated demand for copyrighted materials into dollars.”

Based on all this evidence the Court held that Grokster’s unlawful objective was “unmistakeable.” By contrast, Sony did not try to profit from unlawful taping, it did not intend for its users to unlawfully tape shows, and Sony did not take affirmative steps to encourage its users to use VCRs to infringe copyrights.

The copyright holders wished that the Court had held that P2P technologies were all tools of infringement that had not reached the necessary level of substantial noninfringing uses required under Sony. Grokster, on the other hand, wanted the Court to uphold the Ninth Circuit’s decision that by nature of the decentralized architecture it had successfully avoided having the requisite knowledge of infringing activity required for contributory liability. By deciding the case using an intent-based inducement theory of contributory infringement, the majority opinion avoided this question of when a technology fails to exhibit sufficient potential for lawful uses and therefore imputes on the distributor the direct knowledge of infringement required under the staple article of commerce doctrine. Therefore, the message from the majority seems to be that under the inducement theory you must show by the defendant’s own actions and statements that his unlawful purpose disqualifies him from claiming the protection under Sony. That is, clear evidence of intentional inducement trumps, even in situations where the technology may have substantial noninfringing uses. However, the majority declined to address the conflict about when knowledge of infringement may be imputed to a P2P software distributor saying that it would “leave further consideration of the Sony rule for a day when that may be required.” While all nine justices signed on to the majority opinion, Ginsburg and Breyer both wrote separate concurrences in which they both touched on this issue. In his concurrence, Ginsburg effectively argued for the Seventh Circuit’s approach to determining the knowledge component that would require the court’s to employ a balancing test and would give them more discretion to protect copyright holders. In his concurrence, Breyer basically attacked Ginsburg’s opinion as not doing enough to protect technological innovation. He argued for something closer to the Ninth Circuit’s narrower test.

Ginsburg’s Concurrence

In Ginsburg’s concurrence, he argued that the inducement theory could lead to liability and therefore was an additional tool to protect copyright holders. However, the primary purpose of the concurrence was to say that the Ninth Circuit erred in granting summary judgment to the claim of contributory copyright infringement based on product distribution alone. In addition to addressing whether Grokster was liable under the new inducement theory, Ginsburg believed the court should reconsider whether summary judgment was appropriate for contributory infringement under the traditional test of substantial non-infringing uses under Sony. Under Ginsburg’s approach, liability could be predicated on inducing infringement or simply by distributing a product used to infringe copyrights if the product is not capable of substantial, commercially significant non-infringing uses. She argued that Grokster had not made a sufficient showing of substantial non-infringing use to qualify for summary judgment, saying that there was only “anecdotal evidence” of the P2P technologies being used for non-infringing uses. Therefore, Ginsburg argued for a balancing test more in line with Posner’s recommended balancing test in the Seventh Circuit’s Aimster decision. Ginsburg stated that determining whether there were “substantial” non-infringing uses could only be done by comparing the number or percentage of infringing use to the non-infringing uses.

Breyer’s Concurrence

Breyer’s concurrence, on the other hand, while agreeing with the majority about the potential liability under the inducement theory, disagreed vehemently with Ginsburg’s concurrence saying that the court should reconsider granting summary judgment for contributory infringement under the rule in Sony. Therefore, Breyer thought the Ninth Circuit got it right when it granted summary judgment on the issue of whether Grokster’s product was capable of substantial noninfringing uses. Breyer saw the deeper issue in this case being whether the Court should modify the Sony standard to broaden the scope of contributory copyright infringement claims as Ginsburg suggests. Breyer argued that MGM failed to demonstrate the need to modify Sony and that modifying Sony would impose significant additional risks on technological innovation. Breyer argued that Grokster passes the Sony test because it is “capable of” substantial non-infringing uses. Examples of these non-infringing uses included sharing 1) authorized copies of music from artists like Wilco, Pearl Jam, Dave Matthews, and John Mayer, 2) free electronic books and other works from online publishers, 3) public domain and authorized software such as Winzip, and 4) licensed music videos and television and movies that are distributed via digital video packaging with the permission of the copyright holder. Breyer said it was important that Sony asked whether the product was “capable of” substantial non-infringing uses. The language, he argued, suggested that a court could look at the reasonable prospect of expanded legitimate users over time. Therefore, the tool does not even require substantial non-infringing uses in the shortterm if it appears that there is a reasonable prospect for there to be substantial non-infringing uses in the future. Therefore, Breyer argued that as more and more uncopyrighted material becomes available in digital form, it is reasonable to assume that lawful P2P sharing will become more prevalent. He argued that these non-infringing uses would be likely to increase for sharing things like 1) general research information, 2) public domain films, 3) historical recordings and digital education materials, 4) personal photos, 5) licensed movies and music files, and 6) news broadcasts. For Breyer, determining whether to modify the rule from Sony, as he understands it, requires assessing whether the modification would weaken the protection of technology so much as to outweigh the new copyright-related benefits. He argued that the rule in Sony is very clear and that it rightly recognizes that the goal is not to discourage or control the emergence of new technologies, particularly ones like P2P technology that help disseminate information and ideas more efficiently and broadly. He argued that the “capable of” language in Sony is forward looking and doesn’t confine the analysis of liability to a static interpretation of the product’s existing uses. Therefore, it protects technologies that may have undeveloped future markets. Moreover, he believes the Sony rule is good because it recognizes that judges don’t have specialized technical ability to answer questions about the feasibility or commercial viability of different products. He used the example of filtering technology which Grokster argued would be very difficult and inefficient to create and implement and which MGM claimed would be very easy to do. Under Sony, Breyer believes the judge wouldn’t necessarily have to decide these types of technical questions. Moreover, Breyer argues that using a balancing test like the one proposed in Ginsberg’s concurrence would significantly weaken the law’s ability to protect new technology. To employ Ginsberg’s balancing test, defendants would potentially have to provide detailed evidence of their business plans, profitability estimates, projected feature development cycles and more in order to prove that the benefits of eventual potential lawful uses of the technology outweigh any shortterm issues of copyright infringement. Inventors and entrepreneurs would be subject to potential lawsuits and extensive trials whenever they create and distribute any sort of technology that could be used to infringe copyrights. These inventors would have no way to predict how the courts would weigh the values of infringing versus noninfringing uses. This risk and uncertainty, according to Breyer, would slow down the development of new technologies.

Finally, Breyer argued that in order to decide to modify the Sony rule to better protect copyright holders from the infringements taking place in P2P networks, MGM had to show that the benefits to better protection of copyrights clearly outweighed this weakening of technological innovation. Breyer found that the evidence did not make out a very strong case for changing the rule in Sony. He argued that it was unclear how much revenue to the copyright holders had decreased as a result of the infringement. Furthermore, he argued that part of the role of copyright protection is to ensure that there are adequate incentives to produce creative works and that there was good reason to believe that any decrease in production of creative works under the Sony rule was not substantial. In particular, revenue flowing to musical artists from performances and other sources would remain stable even if the entire CD market was replaced by P2P distribution. He went so far as to say that it would be “silly” to think that music, “a cultural form without which no human society has existed”, will go away because of P2P technology. In fact, Breyer quoted surveys saying that 70% of musicians thought that P2P technology was a minor threat to no threat at all to creative industries. Moreover, Breyer argued that copyright holders had other avenues they could pursue to reduce infringement of their works. For example, they could go after the direct infringers of the copyrights. Breyer quoted statistics showing that when copyright holders pursued these types of lawsuits they had a significant deterring effect. Moreover, the copyright holders could be proactive in developing new technology that would help reduce unlawful infringement. For example, new technologies like “digital watermarking” and “digital fingerpringing” could be used to encode information into files that could help prevent their misuse. Additionally, Breyer quoted evidence that people were willing to pay for P2P style services that sold music over the internet. He argued that many consumers that initially appreciated the convenience of a service like Grokster are now migrating to lawful services in which the cost is worth the benefit of additional convenience and flexibility without the risks of engaging in unlawful infringement. And finally, Breyer argued that courts are less well suited than Congress to attempting to balance the competing interest that inevitably arise with new, emerging technologies.

Posted in Law and Technology | No Comments »

• Pages

  • tmp

• Archives

  • June 2007
  • May 2007
  • January 2007
  • May 2006
  • March 2006
  • November 2005

• Categories

  • Business Process Management (4)
  • Law and Technology (15)

• Meta

  • Login
  • Valid XHTML
  • XFN
  • WordPress