Since speaking last fall on community-based TLDs at the New gTLD Summit in Los Angeles, I have been asked a number of times to provide input on the objections ICANN allows in its New gTLD Applicant Guidebook ("AGB" or simply the "Guidebook"). As the March 13 deadline approaches, I now present the first of a series of four spotlight articles on the subject — one on each of the four permissible grounds for objection.
While many readers have some familiarity with the New gTLD Program and its objection scheme generally, anyone who needs additional information should feel free to contact me directly. Also, if you missed it, listen to ICANN's January 29, 2013 objection webinar here.
Understand the Risks and Benefits Before Filing Objections
Each of my four articles emphasizes a similar idea: the privilege of objecting may be useful, but one must think about it carefully to make it effective. First, the costs — whether fees for filing, attorneys or consultants — are not insubstantial, and will generally exceed what some have come to know in proceedings such as UDRPs. Further, in the case of string confusion, an objector should realize at the outset that if it does not operate an existing TLD, all it can "win" is the right to pull other applicants into a contention set (if not already there).
This first instalment analyzes string confusion ("SC"), the least mechanically complicated of the four. SC has several threshold requirements that any would-be objector must consider before filing, and has separate procedures that come into play once contention sets are finalized.
Objectors Must Meet Strict Standing Requirements
First, a party seeking to object on any ground must have "standing." Using a non-domain example for illustration, someone suing to collect back-payments for veteran's benefits must actually be a veteran or someone acting legitimately on his or her behalf. The Guidebook includes specific standing requirements for each of the four objection types in Section 3.2.2, and potential objectors are well-advised to review them carefully.
With respect to SC, only an "[e]xisting TLD operator or gTLD applicant in current round" may raise an objection. (AGB § 126.96.36.199.) Further, and perhaps even more important, one may bring a SC objection only if ICANN has not already found a likelihood of confusion in its own string similarity evaluation — i.e., if the objector is not already in a contention set. An applicant or operator has to meet both these threshold criteria to be heard by a string confusion panel.
As to contention sets, ICANN announced in its January 11 webinar a March 1, 2013 target date for publishing its string similarity evaluation results. If your string is on that list, don't bother filing a SC objection; it will get thrown out for lack of standing. Better to study up on the string contention procedures in AGB Module 4 and develop a strategy to meet your specific goals.
Logistics of Filing and Planning for Higher Costs
The International Centre for Dispute Resolution ("ICDR") of the American Arbitration Association ("AAA") is the dispute resolution provider for SC objections. For procedural matters, such as where and how to file an objection (and what the filing fees will set you back), the ICDR rules, model forms and other pertinent information can be found here.
Objection documents are filed electronically, as with UDRP actions. Fees to file run from US $8,750 for ex parte proceedings to US $13,000 if a hearing is required. These represent but the first monetary step, however; costs of attorneys or consultants who help would add to that total.
Also keep in mind that New gTLD objections will typically cost significantly more to prosecute than a UDRP, though exactly how much more depends on the type of objection and the specific situation. SC objections present relatively straightforward issues, and more often than not should prove the least costly of the four objection types. Still, a would-be objector should likely budget anywhere from $25,000 to $100,000, depending on the type of objection, just to be safe.
SC Objections Must Satisfy Difficult Standards
If an objector has standing and follows the procedures to file, what must it show in order to be successful? The outcome depends upon the similarity between the objector's string and the one to which it objects — specifically, whether "a string so nearly resembles another that it is likely to deceive or cause confusion in the mind of an average, reasonable Internet user." (AGB § 3.5.1.)
This verbiage appears much like the familiar (to lawyers, anyway) "likelihood of confusion" concept found in trademark law and domain name disputes under the UDRP and many of its ccTLD equivalents. Is it the same? In some ways, yes; in others, no. While it appears to use a number of trademark-related concepts and "buzzwords," the SC standard is, upon close examination, simultaneously broader (in terms of what it covers) and more restrictive (based on what must be shown) than a typical trademark infringement dispute, or even a UDRP action.
Similarities Between SC and Trademark. The "likelihood of confusion" standard requires the SC objector to show it is "probable, not merely possible" that confusion will arise in the mind of the average Internet user. (AGB § 3.5.1.) Such a level of certainty appears in trademark cases as well. See, e.g., Murray v. Cable NBC, 86 F.3d 858, 861 (9th Cir. 1996); WSM, Inc. v. Hilton, 724 F.2d 1320, 1325 (8th Cir. 1984). The easier to prove "likelihood of association" typically does not suffice in trademark cases; nor does it for SC objections: "Mere association, in the sense that [a] string brings another string to mind, is insufficient to find a likelihood of confusion." (AGB § 3.5.1.)
Differences Between SC and Trademark. In some ways, the SC objection cuts a wider swath than trademark. An objector's string need not rise to the level of a protectable trademark to warrant a SC objection. Similarity between the strings is what matters, but not their respective degree of distinctiveness or consumer recognition — the latter being every bit as important as similarity in a typical trademark analysis. However, ICANN has left out some more expansive trademark concepts. "Dilution," for example, involves adverse impact on, rather than likelihood of confusion with, a trademark, and does not even require use in connection with similar goods, services or markets. See, e.g., Perfumebay.com, Inc. v. eBay, Inc., 506 F.3d 1165 (9th Cir. 2007) (allowing dilution claim by online auction provider against cosmetics and perfume retailer). The SC objection language centers solely on potential confusion and the similarity between strings, and does not encompass broader concepts from trademark law such as dilution.
Indeed, ICANN's determination not to include potential dilution-related or "mere association" claims also strongly suggests that it wants SC panels to hold would-be objectors to a high standard when considering questions of similarity. Put another way, ICANN did not intend for the SC objection to serve as another form of legal rights objection, which is a completely separate and distinct claim that does include like additional trademark principles such as dilution that are not raised in the SC context (though more on that in a later instalment). Thus, it is not at all clear, for example, that an objection raised against .STRING would succeed if based on an existing TLD or co-pending application for .STRNG, which could represent typographic variants of other terms such as "strong" or "strung." That .STRNG might "bring to mind" .STRING does not suffice to establish string confusion, AGB § 3.5.1, so the objector would have to show probable confusion between the two separate gTLDs in the minds of reasonable Internet users.
Actual Remedies for String Confusion Are Extremely Narrow
So, if a panel finds a likelihood of confusion, then what? What does the objector win? Well, "win" is a relative term with SC objections. Really, just one of two things can happen.
If the objector's string is an existing TLD, the application with the objected-to string must be rejected outright. (AGB § 188.8.131.52.) If, on the other hand, both strings are in applications in the current round, then the only possible outcome is placing the strings into a contention set with one another. (AGB § 4.1.2.) As a result, a successful objection by a TLD applicant simply places into contention an objected-to string that is not already there by way of ICANN's own string similarity review.
If all this seems a bit anti-climactic, that's because it is. While not totally out of the realm of possibility, the relatively high cost ($8,750-$13,000 plus attorney/professional fees) weighed against the potential relief (not much) makes the SC objection a difficult business proposition. Considering the difficult standard and the potential outcome, it probably would be preferable for applicants to try to reach a "co-existence" relationship, as opposed to filing SC objections against each other.
My next article will examine the limited public interest ("LPI") objection, moving away (temporarily anyway) from the realm of trademark and IP and venturing into some very interesting territory, most notably freedom of speech on the Internet and the problems associated with enforcing legal and societal norms in a global context.
|Data Center||Policy & Regulation|
|DNS Security||Regional Registries|
|Domain Names||Registry Services|
|Intellectual Property||Top-Level Domains|
|Internet of Things||Web|
|Internet Protocol||White Space|
Afilias - Mobile & Web Services
Minds + Machines