Statement
of the
Non-Commercial Users Constituency (NCUC)
on
Domain Name Tasting

7 December 2007

The Final Outcomes Report of the ad hoc group on domain name tasting suggests a growing trend of registrants exploiting ICANN’s Add Grace Period (the “AGP”) to receive a full refund on the cost of registration by canceling their domain name registrations within five days. The AGP may have been adopted upon the assumption that all commercial uses of a domain name would require registration for a period longer than five days. Certain registrants, however, have discovered that they can profit from repeated use of extremely short-term registrations through the use of pay-per-click advertising or otherwise. A coordinated response by ICANN may be appropriate to close this loophole. This response, however, should not be disproportionate to the problem nor stem from any misconception of the issue.

Insofar as some registrants are exploiting the AGP to operate without paying any registrations costs, they are effectively forcing the registries to subsidize them. This was clearly not the intended use of the AGP, and action by ICANN may be appropriate to counter this growing practice. It remains to be seen, however, if the AGP should be removed in its entirely. The ad hoc group report indicates that the AGP may provide benefits to both registrants and registrars, and so completely eliminating the AGP risks eliminating these benefits as well. However, any reported benefits of the AGP are disputed and further elaboration is needed before recommending specific action.

One possible approach may be similar to that adopted by the Public Interest Registries (PIR) — the imposition of a modest ‘excess deletion’ fee. This approach could penalize those registrars with heavy deletions, thus forcing them to adopt policies that prevent registrants from exploiting the AGP. Since registrants looking to avoid paying registration costs will naturally flock to those registrars least vigilant against this abuse, registrars would have a substantial incentive to be vigilant against creative disguises of these practices. Yet unlike directly imposing a fee on all short-term registrations, this approach gives registrars significant flexibility to adopt effective practices tailored to their customer base and business model, and preserves the other advantages of the AGP.

Intellectual Property Issues

The intellectual property issues discussed in the ad-hoc group’s final report warrant special attention. In this context, “intellectual property” refers almost exclusively to trade and service marks, which are often referred to collectively as “trademarks.” The vast majority of the respondents to the RFI identified themselves as either intellectual property rights owners (37.93%) or representatives of intellectual property rights owners (51.23%). Consequently, intellectual property rights feature prominently in the responses.

The problem which domain tasting presents to trademark holders is not that the AGP creates a loophole which makes otherwise infringing activity legal. If a registrant makes use of a trademark in a manner that constitutes infringement, the holder of that trademark is protected through international treaty, the laws of various nations, and through ICANN’s own Uniform Dispute Resolution Policy. These protections still apply even if the period of registration is very brief. The problem is instead one of enforcement.

This distinction should be kept in mind by the GNSO and by any subsequent working group established to tackle this issue. Many of the responses to the RFI listed problems such as “erosion of brand names,” “erosion of reputation” and “loss of revenues [through] diversion of traffic” as disadvantages to domain tasting. These are problems with infringement, not with domain tasting. While it may be appropriate for ICANN to consider whether its policies unduly encourage infringement or impede enforcement of intellectual property rights, it would be a mistake to assume that a revised policy on domain tasting will stamp out short term infringement or that all domain tasting necessarily infringes.

Insofar as the AGP allows a registrant to use a domain for a very short time at no cost it does provide an incentive to a prospective infringer to operate in a manner that frustrates enforcement of trademark rights. This incentive can be removed by implementing a modest restocking fee where no corrective motive can be shown for the deletion. Because the bulk of deletions come from a handful of registrars and because registration fees are only likely to deter an infringer who operates a large number of sites, the approach adopted by PIR (option “C” on the RFI), is particularly worth further consideration.

The Sample Zone File Data Study

ICANN should be particularly careful in crafting any test to identify infringing activity. One proposal in the ad hoc group’s report was to determine the percentage of domain tasting that infringed upon trademarks by comparing a sample of deletions to a list of trademarks registered with the United States Patent and Trademark Office (the “USPTO”). This method was termed the “sample zone file data study.” This method would result in erroneous and excessive findings of infringement because it stems from a fundamental misconception of trademark law. Specifically, it relies upon an erroneous assumption that any unauthorized use of a registered trademark is unlawful.

Trademark law does not categorically ban use of a trademark without the permission of the owner. Instead, it prohibits uses of a trademark which deceive or confuse the consumer. Where there is no confusion, there is no infringement. Thus, trademark law does not prohibit the use of the same name or symbol by companies in different fields of commerce, and is limited in terms of its geographical reach. Therefore a test for infringement based solely on the presence of a word that has been registered with any trademark office would erroneously conclude that many lawful business uses are infringing.

This is easily illustrated by examining one registered trademark. The USPTO lists 125 live registered wordmarks which contain the word “Acme.” Many of these are simply the word “Acme” with little or no graphical embellishment. Yet hundreds of Corporations, Limited Partnerships, and Limited Liability Companies with names containing the word “Acme” have been registered with the California Secretary of State, to say nothing of General Partnerships or unincorporated Sole Proprietorships in California or business entities in other jurisdictions. While a few of these businesses may be infringing upon the trademarks of others, the vast majority are undoubtedly operating without any consumer confusion. Moreover, it may be possible to start a new business incorporating the word “Acme” without infringing upon any of those trademarks registered. Under the sample zone data file study, however, any domain incorporating the word “Acme” would be inferred to be infringing merely because this word has been registered with the USPTO.

More significantly, non-commercial uses of a registered trademark would also be determined to be infringing under the test proposed. Under U.S. Law, non-commercial use is particularly unlikely to be found to infringe because there is little chance of confusion. Thus a website critical of Jerry Falwell which used a common misspelling of his domain name (“Fallwell.com” for “Falwell.com”) was ruled to not infringe upon his trademark because the creator intended “only to provide a forum to criticize ideas, not to steal customers.” Since on-line critics of businesses frequently incorporate the name of the criticized business into their domain names (e.g. “paypalsucks.com,” “microsoftsucks.org,” etc.) false findings of infringement are particularly likely under the sample zone file data study discussed in the report.

To be sure, an argument can be made that non-infringing domains are less likely to be deleted during the AGP. If that is the case, then it is less likely that these legal uses of registered trademarks would significantly skew the sample zone file data study’s conclusions. It would be a mistake, however, to use that argument to justify the proposed test. This test is intended to determine whether infringing use predominates in the practice of domain tasting. To argue that a use of a trademark is probably infringing because it is deleted during the AGP is to assume the outcome the test is intended to determine—a logical fallacy known as “begging the question.”

More importantly, ICANN should be careful not to establish a precedent that this fundamentally flawed test establishes infringement. Given the difficulties inherent in enforcing trademark rights against domain tasters, it is possibly that some sort of mechanism to screen-out infringing use will be discussed during the policy development process. The test proposed for the sample zone file data study would be manifestly inadequate for this purpose in that it would prevent a great deal of legitimate use.

This last point is particularly significant in light of the fact that trademark law is still adapting to commerce over the Internet. For example, while some U.S. Courts have held that a bad faith intent to make money from a domain containing a famous trademark is sufficient to establish infringement, others have held that such a use must be in connection with some form of goods or service. ICANN should not take it upon itself to decide these issues for the courts and legislatures of every country. The delicate balance of competing public policies inherent in intellectual property law should instead be left to the courts and political processes to work out.

Conclusion

Further investigation within the GNSO is needed and action may be required to curb abusive domain name tasting. As the GNSO takes the next step in dealing with this problem it must be careful to ensure that the issue remains properly framed rather than assuming than ICANN is responsible for or capable of preventing all short-term trademark infringement on the web. Moreover, while further investigation, discussion, and action is warranted at this point, the proposed sample zone file data study should not be undertaken because it relies on a fundamental misunderstanding of trademark law and sets a dangerous precedent as to what ICANN will consider to be infringing use.