Constructive knowledge is a flexible rule, not a bright-line standard. That induces enforcement flexibility, cooperation between brands and platforms, and prevents under-deterrence. Platforms can discharge liability through proactive enforcement
David Bernstein & Michael Potenza, 2011 - *partner with Debevoise & Plimpton LLP in New York, and an adjunct professor of law at New York University Law School and George Washington University Law School, where he teaches Advanced Trademark Law AND **counsel with Debevoise & Plimpton LLP. “Why the Reasonable Anticipation Standard Is the Reasonable Way to Assess Contributory Trademark Liability in the Online Marketplace,” 2011 STAN. TECH. L. REV. 9 (2011). Hein Online, accessed via University of Michigan //DH
The common law principles discussed above are based on the bedrock concept of reasonableness under the circumstances. Properly applied, these principles contemplate precisely the paradigm of joint cooperation that best furthers trademark policies. With that in mind, an appropriate standard for contributory liability in the online marketplace should contain the following features. First, it should provide for liability based on actual knowledge of infringing conduct by third parties, coupled with a failure by the online marketplace to remove the infringing listings. Such a test would cover the active inducement and actual knowledge standards recognized in Ives. Such a provision should also include an explicit recognition that actual knowledge may be satisfied with willful blindness, which has long been recognized as being equivalent to knowledge, even in trademark cases.147
Second, the proper standard should provide for liability through constructive knowledge, which can be established where the online marketplace fails to take reasonable precautions in circumstances in which the infringing conduct can be reasonably anticipated. This prong is based on the common law standards embodied in the language of the Restatement (Third) of Unfair Competition § 27. As discussed in Part III, the "reasonably anticipated" language is based on the "has reason to know" standard in tort law. Thus, under this test, trademark-infringing activity can be "reasonably anticipated" when a reasonable marketplace operator under like circumstances would have reason to know of infringing third-party conduct. As discussed in Part III above, a consideration in determining whether infringing third-party activity can be "reasonably anticipated" is the prevalence of third-party infringing activity in the marketplace. As in Snow Crest and Ives, the plaintiff would bear the burden of demonstrating prevalence sufficient to trigger reasonable anticipation.
It should be emphasized that an online marketplace that can reasonably anticipate subsequent infringing activities by others "is required to take only those precautions that are reasonable under the circumstances."148 An online marketplace that has taken "reasonable precautions" would not be liable even if third parties continue to engage in infringing or counterfeit transactions. 149 A similar two-prong test providing for liability based either on actual knowledge or on constructive knowledge measured by a reasonableness standard has been adopted in the European Union. The European Court of justice held that an online marketplace may be found liable where it has "actual knowledge of illegal activity" by its users and has not "acted expeditiously to remove or disable access to, the information."'150 Under this prong, actual knowledge can be inferred where the online marketplace has "provided assistance" to third-party infringers (e.g., by "optimising the presentation of the offers for sale ... or promoting those offers").151 Even if the online marketplace has no actual knowledge of infringing activity, it may nonetheless be held liable if it is "aware of facts or circumstances on the basis of which a diligent economic operator should have identified the illegality in question and acted." 152 Awareness of facts or circumstances that trigger this duty to act may arise from, for example, the online marketplace's own investigation or notifications from brand owners or other users.153
In addition to remaining consistent with common law traditions and cases such as Snow Crest, a standard featuring these two prongs would be flexible enough to advance the policy considerations discussed above. In particular, the "circumstances in which the counterfeiting or trademark infringing activity could be reasonably anticipated" will vary based on the practical and technological considerations that apply to a particular marketplace, and the corresponding "reasonable precautions" will also depend on the circumstances. In this way, the standard does not set a bright line rule that always places most of the burden on a particular party, but rather allows that burden to be properly shared. The standard does this by triggering a duty on the marketplace only to take precautions that are commensurate with the scope of the infringement problem.
For example, in the case of an online marketplace where virtually all Tiffany goods sold are counterfeit, the only "reasonable precaution" may be for the online marketplace to place a blanket prohibition on all Tiffany goods unless their authenticity is established. In such a case, the legal standard set out above would place most, if not all of, the burden to prevent infringement on the marketplace owners. In the words of Snow Crest, where "most" of the transactions infringe, the general knowledge of that prevalent activity triggers the reasonable precaution duty.154 In other circumstances, where counterfeiting does not post as stark or obvious a risk, a more even sharing of the burden would be appropriate. For example, where many articles are noninfringing, it may make more sense (as Snow Crest recognizes) to require specific knowledge on the part of the online marketplace. 55 The "notice and take-down procedures"'16 employed by online marketplaces like eBay, therefore, may be a "reasonable precaution" that would counsel against liability where there are many perfectly legitimate transactions involving a particular brand. This recognizes that trademark owners, working cooperatively with online marketplace providers, are the most efficient cost avoider. Another example of this type of burden-sharing "reasonable precaution" would be online marketplaces providing space for trademark owners to publish a page educating buyers about their products with instructions on how to detect fakes.
At the same time, the standard must be flexible enough to evolve with changes in technology. For example, if new technologies provide ways to determine the authenticity of an item by reviewing a scan of a photograph of the item, that might impose a duty on the online marketplace to adopt such technologies. Such a standard might also provide an incentive for brand owners to invest in the development of such technologies, or to include secret markings on their goods to facilitate such scanning and identification of authenticity.
The central point is not that courts should develop a rigid checklist of what precautions are reasonable, but that an appropriate standard must allow for a fact-intensive inquiry that does not limit liability to the knowledge of specific instances of infringing conduct. The trier of fact must be able to consider whether, under all the circumstances, general knowledge of widespread infringing conduct is sufficient to impose a duty to take further reasonable measures. If circumstances triggering a duty to take reasonable precautions are found to exist, the marketplace owners may discharge their duty by taking anti-counterfeiting measures commensurate with the scope of the problem. These may include electronic measures to filter out new listings or identify existing listings that are offering infringing goods, or other systems for identifying, suspending, and banning infringing users by IP address. The European Court of Justice has also identified the suspension of third-party infringers and the use of measures that make it easier to identify an online marketplace's customer-sellers as two examples of precautions that a court might, under appropriate circumstances, require an online marketplace to adopt.157
Applying such a standard to the dispute between eBay and Tiffany, it would be up to a trier of fact to determine (1) whether 75% of Tiffany items sold on eBay in 2004 and 2005 were indeed counterfeit, as Tiffany argued; 158 (2) whether eBay knew of this fact; (3) whether this knowledge, though general, was sufficient to trigger "reasonable anticipation;" and (4) whether eBay's precautions like its "fraud engine"159 and VeRO Program,"160 constitute reasonable precautions in light of the totality of the circumstances. The standard should be flexible and should appropriately focus a trier of fact on these issues, not whether eBay had specific knowledge of particular infringing listings. Based on Judge Sullivan's ruling, it appears that the court would have found that eBay was not contributory liable. Had the court applied this more nuanced, balanced standard, it would have created more efficient incentives, more fully supported the policies underlying trademark and tort law and promoted more equitable results in future cases.
Although the standard proposed in this paper is put forth specifically with the online marketplace in mind, it bears emphasis that the standard is derived from, and remains equally applicable to, the brick-and-mortar context. Indeed, the fact that the proposed standard has its origins in the common law emphasizes its implicit position that there is nothing unique about the Internet that justifies a different standard. In Lockheed Marin Cop. v. Network Solutions, Inc., the court, after considering contributory trademark infringement in the flea market context, held that the application of the Ives standard does not turn on whether the service provider is online or in brick-and-mortar. 161 Rather, "[d]irect control and monitoring of the instrumentality used by a third party to infringe the plaintiffs mark permits the expansion of Ives's 'supplies a product' requirement for contributory infringement" to a service provider.162 The district court in Tiffany applied the Lockheed Martin analysis and determined that "eBay exercises sufficient control and monitoring over its website such that it fits squarely within the Fonomsa and Hard Rock Cafe line of [flea market] cases," and therefore the same standards should apply.163
Certainly, online marketplaces differ in significant ways from brick-and-mortar ones. Courts considering contributory liability and reasonable precautions in an online context should consider those different facts. Reasonable precautions in the case of eBay may quite possibly be satisfied by its existing measures such as its fraud engine and VeRO program, whereas a reasonable precaution in the brick-and-mortar flea market context may involve the market owner walking up and down the aisles of the market looking for counterfeit goods. In addition, it is certainly possible that different business models within a single online marketplace system would cause courts to conclude that different reasonable precautions are necessary. This is no different from how the courts apply the common law standard differently in the brick-and-mortar context to a single retail location as compared with a multi-acre swap meet.
CONCLUSION
As e-commerce continues to explode and online marketplaces come to play a larger and more prominent role in global trade, the opportunities for trademark counterfeiters to leverage the Internet's anonymity to free-ride off the goodwill of strong and reputable marks will also expand. The resulting erosion of the source-identifying capacity of trademarks will result in an increase in consumer search costs and economic harm to brand owners and to online marketplaces. The prevalence of counterfeiting and infringement will also diminish consumer confidence in online marketplaces as trustworthy alternatives to traditional brick-and-mortar markets.
The Second Circuit's decision in Tiffany, by effectively reading the "reason to know" standard out of Ives and imposing a rigid "specific knowledge test," robs the common law doctrine of its essential flexibility. The Tiffany rule effectively places the entire burden of trademark enforcement on the owners of marks and will under-deter the incident of trademark infringement and counterfeiting.
An appropriate legal standard should recognize that generalized knowledge is, as it always has been, probative of whether the online marketplace has "reason to know" of the direct infringement and therefore may have a duty to the brand owner to take reasonable precautions. Whether this duty is triggered, as Snow Crest recognizes, is a fact-sensitive inquiry that, at bottom, depends on how prevalent the underlying tortious activity is. This kind of standard acknowledges that both online marketplaces and trademark owners have a role to play in the protection of trademark rights online. Recognizing this type of standard would restore uniformity to this "ill-defined" area of law and provide clear guidance to brand owners and online marketplaces alike to ensure the legitimacy of transactions in this increasingly vital sector of the national and global economy.