IRT Report overreaches its mandate
Electronic Frontiers Australia is an online civil liberties group based in Australia. We have concerns with the IRT as presented, and regard several of its proposals as poorly considered and unbalanced. The IP Clearing House system is, in theory, a potentially useful adjunct to registries that wish to use it. In practice we remain unconvinced that a single Clearing House would be practical for all gTLDs, regard the proposal as outside ICANNs scope, and the proposal as discussed skips over a large number of practical issues that would make a single, ICANN mandated, Clearing House problematic. The Globally Protected Marks list we regard as one of the most problematic and unnecessary aspects of the IRT. It creates an entirely new class of mark, unsupported in common trademark law, with special administrative rights unrelated to existing legal rights. And the GPML appears to be a complex solution to a relative non-problem. Deceitful registration of a gTLD would appear to be an issue already adequately dealt with by civil remedies or by extending the UDRP or a more complex dispute resolution fee. And in the cases where a well known trademark happens to also have other, legitimate, applications it is unclear why there is a need for certain (and only some) trademark holders to be given special consideration by ICANN. GTLDs, with $185,000 filing fees, are unlikely to be plagued by simple cyber-squatting, and special protection at the gTLD level seems to be designed to solve a problem that is unlikely to occur, and for which adequate remedies exist. As far as second level domain rights protection mechanisms, it seems clear that a single set of rules governing trademark protection will not be appropriate for all gTLDs. Any one-size-fits-all approach, such as suggested by the IRT, will be inappropriate for many, if not most, gTLDs. It may be useful for the IRT to suggest a model approach, but it seems almost certain that it will be need to be tailored to the specific needs of a gTLD. For example, trademarks apply only to certain types of goods and services, and many gTLDs may clearly indicate a particular range of goods and services, thus making many trademarks irrelevant. For example, within the proposed .food gTLD, the goods and services discussed are clearly those relevant to food - there is no reason why even famous trademark holders whose trademarks apply to goods and services unrelated to food (such as Apple computer, and Apple Records) should even be considered when the gTLD itself indicates that apple.food would likely apply to the foodstuff. On the issue of the URS, EFA understands that the UDRP can be a problematic and inadequate response to some patterns of abusive registration. But the URS as proposed has many problems. It needs to be far better integrated with the existing UDRP system, so that it complements the existing process. It should always be possible for the target of a URS to essentially assert their belief that a site is not infringing (or at least, not trivially so) and demand the UDRP process. A proposed fast resolution system like the URS should not replace the UDRP. There are also issues with the timeline and processes of the URS as described. And also, the URS process as described consider only web traffic - this is clearly inadequate and inappropriate and needs reconsideration. Many domain names are still used primarily for purposes other than web traffic. Furthermore, once again the assumption is made that a single administrative procedure is appropriate for all gTLDs. For many gTLDs a process like the URS may be sufficient - for many others, it may clearly be inappropriate, such as potential gTLDs set up for free speech purposes. The URS may well form the proposed basis for a model procedure, but may still be inappropriate for some gTLDs. There are also some doubts about the practicality of the URS. In our opinion a URS type scheme should only be introduced as part of a broader process involving a range of different malicious behaviour, not just trademark violation (spam and phishing, for example), and preferably considered at the same time as a review of the entire UDRP process. The Thick Whois model is extremely problematic from a privacy rights point of view. If followed, it would lead registries into a legal minefield as regards EU privacy legislation and other national level privacy laws. It may be appropriate for some registries and some gTLDs, but is likely to be extremely inappropriate for others. Once again, the Thick Whois model may be best practice for some domains, but is quite likely to be highly inappropriate for others, particularly where free speech considerations are an issue. In conclusion, the IRT report puts forward many proposals that may be appropriate for many commercial domains, but presents a simplistic and one-size-fits-all view of the trademark world and the many uses to which gTLDs might be put. As a set of suggestions as to how new gTLDs might best implement trademark protection, the IRT proposals contain some ideas that deserve serious consideration and may form the basis for a model proposal, but very few of the IRT proposals should be globally mandated on new gTLDs.
Regards David Cake Secretary, Electronic Frontiers Australia