This comment is being presented in my own personal capacity and does not represent the views of my employer, NeuStar, Inc., and its subsidiaries or affiliates, or the Implementation Recommendations Team.

Preamble

I recently learned about a meeting that took place between ICANN staff and Noncommercial Users Constituency (NCUC) members Kathy Kleiman and Konstantinos Komaitis regarding the Implementation Recommendations Team (IRT) recommendations for the protection of intellectual property rights in new generic Top-Level Domains (gTLDs). My comment relates to the White Paper published by Ms. Kleiman and Mr. Komaitis (hereinafter referred to as the “KK Proposal”), posted here, with respect to the notion of having multiple Regional Trademark clearinghouses (TMCs).

For the reasons stated in this comment, the KK Proposal fails a number of the benchmarking checklists used by the IRT in evaluating proposals. More specifically, I believe that the proposed multiple clearinghouse model: (a) does not adequately address the harms against which the clearinghouse was set up to protect; (b) will not scale; (c) can be gamed and abused; (d) is not the least burdensome solution; and (e) is not technologically feasible at this stage if a new TLD process is to be launched in 2010.

Of particular interest is the e-mail that was sent by Ms. Kleiman to the NCUC describing the meeting she had with ICANN staff. She notes:

Konstantinos and I wanted to share with ICANN some of our views as experts in the field who have analyzed the IRT Report closely and consulted widely with NCUC members, ALAC, Registrars, Registrant Attorneys and IRT members.

Note that Ms. Kleiman does not reveal the names of the persons or entities with whom she consulted, and that existing and future gTLD registries—a group that I regard as highly important and very germane to this discussion—is completely ignored. It is unfathomable to me that any “innovative solutions” for the protection of intellectual property rights could possibly be developed by a group of people that does not include any representatives of the very organizations that will be responsible for implementing such solutions.

Although some may not agree with the recommendations the IRT has made, I must credit them for allowing the participation of a registry and registrar to discuss solutions that were actually implementable. As I suggested to the IRT during our meetings, I believe that having multiple clearinghouses is not implementable in the immediate future—and would require a significant delay to the new gTLD process, which is something I personally would not like to see.

Introduction

According to the IRT Recommendations, the IP clearinghouse is intended to be used for a number of rights protection mechanisms, including: (a) the validation of trademark rights on an annual basis; (b) a globally protected marks list; (c) a pre-launch IP claims service; and (d) the generation of data for and participation in URS pre-registration, and validation of URS complaint claims regarding trademark rights. In addition, although it is not mentioned in the KK Proposal, one of the key functions of the IP clearinghouse is to consult with new gTLD registry operators (at no charge) to assist those registry operators in drafting policies and creating processes to implement the RPMs. Having multiple regional TMCs will not only hinder efforts to achieve the latter goal, but will in many ways render the entire objective unattainable. More specifically:

a) The KK Proposal is much more complex than the IRT Proposal. The KK Proposal calls for market-based regional clearinghouses to “self-form” and provide these functions with ICANN “encouragement,” but not ICANN oversight. How will these clearinghouses form? Who will approve or accredit these entities, and how? What are the criteria to approve all of these entities? What if there are no clearinghouses that form in any of the ICANN regions? Who will serve them?

b) The KK Proposal is inconsistent. The KK Proposal states that ICANN does not have the “scope nor authority” to create one single clearinghouse—yet, somehow, the authors of the KK Proposal want readers to believe they have the authority to approve multiple clearinghouses.

c) The KK Proposal will dramatically increase costs and inefficiencies, contrary to statements made in the proposal. The KK Proposal correctly recites the fact that the clearinghouse is meant to reduce costs and be efficient—but then goes on to state that “it fails, since it suggests a structure that depends much on ICANN, registries and registrars.” The KK Proposal does not go into any other detail as to why this is an issue, and as such, this assertion makes no sense to me. Any proposal to deal with rights protection mechanisms must involve ICANN, the registries and registrars; after all, that’s the whole point of the clearinghouse.

To state the obvious: Registrars accept domain name registrations from registrants and submit them to the registries, and this procedure remains consistent throughout every Sunrise and/or IP Claims process. To be incredibly simplistic even further, the registries are responsible for ensuring that any registrations they accept from registrars during a Sunrise process are validated. Only upon validation can an application made during a Sunrise process be “registered.” The registrar is responsible for collecting all information about the registrant necessary for obtaining a domain name registration. Thus, any rights protection mechanism must involve—if not depend on—the registries and registrars, because those parties are the ones that will be required to implement the solutions.

I completely disagree with the KK Proposal’s assertion that it will be more cost-effective and efficient to have multiple TMCs. Each of the registries that have implemented Sunrise protection mechanisms have, in fact, relied on a single validating body. In fact, no ccTLD or gTLD that has implemented a Sunrise process to date (including .asia, .eu, .me, .mobi, .tel and .us, to name a few) has ever used multiple validators. This is because it is neither a trivial nor an inexpensive task to build an automated process to exchange information between a validator and the registry. The complexity of that task will be greatly increased if you have multiple TMCs speaking multiple languages in multiple regions and in multiple time zones.
What may not be known or recognized by the authors of the KK Proposal is that for each of the past Sunrise processes, the registries, registrars and validators have been in constant communication. This not only involves dealing in the design and implementation of Sunrise processes, but also in customer support, technical operations and troubleshooting. These processes are all tailored to the individual registry and the specific validator. Having one clearinghouse that serves as the validator adds predictability, uniformity and reliability in the design and implementation of these processes. If a gTLD registry operator is forced to deal with multiple TMCs, the performing of these functions would become exponentially more difficult. Specifically, each registry would have to separately deal with each individual TMC to make sure that its proposed unique policies are acceptable and implementable by each and every TMC. This would result in huge costs to the registry—and ultimately to the intellectual property owners and registrants.

d) The KK Proposal results in lack of uniformity. The KK Proposal advocates that each TMC “foster a common database search mechanisms for registrars to search trademarks by country, date of registration, international classification . . .” Aside from not recognizing that it is the registries (and not registrars) that are responsible for the validation of Sunrise registrations, it fails to recognize the whole point of having the clearinghouse in the first place. The clearinghouse is being established to ensure that an intellectual property owner has to have its rights validated only once despite the potential of hundreds of new TLDs. In the past, each registry set up its own Sunrise process, and each process was very different. The IRT wanted to take this job away from each domain name registry and place it in the hands of one entity. The one clearinghouse would then be responsible for answering the simple question: “Does the purported intellectual property rights owner have the rights it claims to have (in a given class and jurisdiction), and does that meet the criteria set up by the registry?” If the answer to these questions is yes, then the registration proceeds. In the case of multiple TMCs, you have the strong possibility of different interpretation of those questions, resulting in different answers—which, of course, completely goes against the objective set by intellectual property owners, registrants and the registries.

e) The KK Proposal will result in Forum Shopping. In addition to the above, I believe that instituting a multiple TMC model will result in “forum shopping” by intellectual property owners. IP owners with multiple registrations in multiple regions would simply choose the most favorable TMC to use as its validator. (Put differently, if each region uses its own rules to verify the supporting documentation establishing trademark rights, logic dictates that IP owners could certainly opt to use the TMC in the region that is the most favorable to them.) On the other hand, having one IP clearinghouse does not preclude the consideration of jurisdictional nuances, but rather better ensures consistent review with jurisdictional nuances also being considered. With standard review comes increased assurance that trademark rights are not unwittingly expended.

f) The KK Proposal is inconsistent on the issue on “Unregistered Marks.” The KK Proposal strongly advocates against allowing unregistered trademarks to be used as a basis for a Sunrise Process. The proposal states that “the validation concerning established legal rights falls outside the scope of ICANN.” Ironically, the KK proposal also cites the .eu sunrise and PriceWaterhouseCoopers as a model for future Trademark clearinghouse—completely glossing over the fact that the .eu Sunrise was in fact the first (and one of the only) Sunrise Processes to involve the very common law rights that the KK Proposal so vehemently argues against using. For more information, see Slide 11 of the presentation [PDF].

Contrary to the KK Proposal’s assertion, and as repeatedly stated by the IRT in each consultation, the IRT is not advocating that all Sunrise periods include processes that validate common law or unregistered rights. The IRT recommended that the IP clearinghouse have the ability to collect information about unregistered rights if (and only if) a new gTLD registry wanted, for whatever reason, to have a policy that allowed the use of common law or unregistered rights. Remember, the IP clearinghouse is not intended to create any policy or any legal rights that do not presently exist. Rather, it is the job of the new gTLD registry to set the policy with respect to rights protection mechanisms (subject to the minimums established by ICANN), and it is the job of the validator to confirm the existence of supporting documentation that establishes their claim of rights, certification or existence. For example, if a registry for .pizza wants to allow any owner of a pizza shop to have preferential treatment in a Sunrise process absent a trademark registration, then the IP clearinghouse should have the ability to collect that information and verify it.

g) No gTLD Sunrise process to date has used or even needed multiple clearinghouses. The KK Proposal implies that the use of multiple clearinghouses is necessary to conduct a fair, cost-efficient, and proper Sunrise Process. What the KK Proposal ignores is the fact that there have been a number of gTLDs (and ccTLDs) that have already launched using one clearinghouse. I am not aware of ANY complaints that have been lodged against any of these Sunrise processes due to the fact that only one clearinghouse was used. In addition, there has never been a complaint (as the KK Proposal would have you believe), that having one clearinghouse provided has lead to “increased costs (administrative and external)” or that it “led to litigation and create[d] bureaucracy and confusion.” Moreover, there has not been any complaint that the one clearinghouse was unable to deal with trademark registrations around the world. To the contrary, the Sunrise Processes employed in .asia, .mobi, .eu., .me and .tel (as just a few examples), all using one clearinghouse, have accepted trademark registrations from countries in every region of the world. After having all of these Sunrise Processes, where not one complaint was lodged for having only having one clearinghouse, why is it now necessary to have multiple clearinghouses? In addition, isn’t it more plausible that having multiple clearinghouses would lead to increased costs, bureaucracy, confusion and litigation?

h) At this point in time, having multiple TMCs would significantly delay the introduction of new gTLDs. As I explained to Mr. Komaitis at the London consultation, having multiple TMCs could result in a significant delay in the introduction of the new gTLD process. This is because it may necessitate the development of a protocol by which new gTLD registries would have to have a mechanism to communicate with multiple clearinghouse providers—or, alternatively, a registry of clearinghouse providers—to query the clearinghouse providers to determine which provider actually has the information of the specific intellectual property owner submitting the Sunrise registration, and then to query the appropriate clearinghouse provider to have that information validated once it is determined. As you may be aware, registries and registrars are required to use established protocols to communicate with each other. These protocols go through multiple-year processes with the Internet Engineering Task Force (IETF). The current communication protocol (EPP) took a number of years to work its way through the complex IETF standards process. I am uncertain that the communication required under the KK Proposal can be accomplished through EPP—and if it does not, it would have to go through the formal standards process prior to approval.

Even if no new protocols would have to be developed, a new standard API still would have to be developed in order for each of the registries to use them with each of the TMCs. Getting multiple TMCs together to design such a common API (once ICANN approves them) could take a significant amount of time. On the other hand, having one clearinghouse that can dictate the specifications of the API would reduce time-to-market considerably.

In conclusion, a “one clearinghouse” model does not preclude all of the considerations that the KK Proposal is worried about—and as I have demonstrated, has more of an ability to address the concerns in a consistent manner to avoid an expansion of rights and unnecessary delay in the implementation of new gTLDs.