Archives For Top-level domain

One of the main concerns I had during the IANA transition was the extent to which the newly independent organization would be able to behave impartially, implementing its own policies and bylaws in an objective and non-discriminatory manner, and not be unduly influenced by specific  “stakeholders”. Chief among my concerns at the time was the extent to which an independent ICANN would be able to resist the influence of governments: when a powerful government leaned on ICANN’s board, would it be able to adhere to its own policies and follow the process the larger multistakeholder community put in place?

It seems my concern was not unfounded. Amazon, Inc. has been in a long running struggle with the countries of the Amazonian Basin in South America over the use of the generic top-level domain (gTLD) .amazon. In 2014, the ICANN board (which was still nominally under the control of the US’s NTIA) uncritically accepted the nonbinding advice of the Government Advisory Committee (“GAC”) and denied Amazon Inc.’s application for .amazon. In 2017, an Independent Review Process panel reversed the board decision, because

[the board] failed in its duty to explain and give adequate reasons for its decision, beyond merely citing to its reliance on the GAC advice and the presumption, albeit a strong presumption, that it was based on valid and legitimate public policy concerns.  

Accordingly the board was directed to reconsider the .amazon petition and

make an objective and independent judgment regarding whether there are, in fact, well-founded, merits-based public policy reasons for denying Amazon’s applications.

In the two years since that decision, a number of proposals were discussed between Amazon Inc. and the Amazonian countries as they sought to reach a mutually agreeable resolution to the dispute, none of which were successful. In March of this year, the board acknowledged the failed negotiations and announced that the parties had four more weeks to try again and if no agreement were reached in that time, permitted Amazon Inc. to submit a proposal that would handle the Amazonian countries’ cultural protection concerns.

Predictably, that time elapsed and Amazon, Inc. submitted its proposal, which includes a public interest commitment that would allow the Amazonian countries access to certain second level domains under .amazon for cultural and noncommercial use. For example, Brazil could use a domain such as http://www.br.amazon to showcase the culturally relevant features of the portion of the Amazonian river that flows through its borders.

Prime facie, this seems like a reasonable way to ensure that the cultural interests of those living in the Amazonian region are adequately protected. Moreover, in its first stated objection to Amazon, Inc. having control of the gTLD, the GAC indicated that this was its concern:

[g]ranting exclusive rights to this specific gTLD to a private company would prevent the use of  this domain for purposes of public interest related to the protection, promotion and awareness raising on issues related to the Amazon biome. It would also hinder the possibility of use of this domain to congregate web pages related to the population inhabiting that geographical region.

Yet Amazon, Inc.’s proposal to protect just these interests was rejected by the Amazonian countries’ governments. The counteroffer from those governments was that they be permitted to co-own and administer the gTLD, that their governance interest be constituted in a steering committee on which Amazon, Inc. be given only a 1/9th vote, that they be permitted a much broader use of the gTLD generally and, judging by the conspicuous lack of language limiting use to noncommercial purposes, that they have the ability to use the gTLD for commercial purposes.

This last point certainly must be a nonstarter. Amazon, Inc.’s use of .amazon is naturally going to be commercial in nature. If eight other “co-owners” were permitted a backdoor to using the ‘.amazon’ name in commerce, trademark dilution seems like a predictable, if not inevitable, result. Moreover, the entire point of allowing brand gTLDs is to help responsible brand managers ensure that consumers are receiving the goods and services they expect on the Internet. Commercial use by the Amazonian countries could easily lead to a situation where merchants selling goods of unknown quality are able to mislead consumers by free riding on Amazon, Inc.’s name recognition.

This is a big moment for Internet governance

Theoretically, the ICANN board could decide this matter as early as this week — but apparently it has opted to treat this meeting as merely an opportunity for more discussion. That the board would consider not following through on its statement in March that it would finally put this issue to rest is not an auspicious sign that the board intends to take its independence seriously.

An independent ICANN must be able to stand up to powerful special interests when it comes to following its own rules and bylaws. This is the very concern that most troubled me before the NTIA cut the organization loose. Introducing more delay suggests that the board lacks the courage of its convictions. The Amazonian countries may end up irritated with the ICANN board, but ICANN is either an independent organization or its not.

Amazon, Inc. followed the prescribed procedures from the beginning; there is simply no good reason to draw out this process any further. The real fear here, I suspect, is that the board knows that this is a straightforward trademark case and is holding out hope that the Amazonian countries will make the necessary concessions that will satisfy Amazon, Inc. After seven years of this process, somehow I suspect that this is not likely and the board simply needs to make a decision on the proposals as submitted.

The truth is that these countries never even applied for use of the gTLD in the first place; they only became interested in the use of the domain once Amazon, Inc. expressed interest. All along, these countries maintained that they merely wanted to protect the cultural heritage of the region — surely a fine goal. Yet, when pressed to the edge of the timeline on the process, they produce a proposal that would theoretically permit them to operate commercial domains.

This is a test for ICANN’s board. If it doesn’t want to risk offending powerful parties, it shouldn’t open up the DNS to gTLDs because, inevitably, there will exist aggrieved parties that cannot be satisfied. Amazon, Inc. has submitted a solid proposal that allows it to protect both its own valid trademark interests in its brand as well as the cultural interests of the Amazonian countries. The board should vote on the proposal this week and stop delaying this process any further.

As we noted in our issue brief on the impending ICANN transition, given the vast scope of the problem, voluntary relationships between registries, registrars and private industry will be a critical aspect of controlling online piracy. Last week the MPAA and registry operator Radix announced a new “trusted notifier” program under which the MPAA will be permitted to submit evidence of large-scale piracy occurring in Radix-managed top-level domains.

In many respects, this resembles the program that the MPAA and Donuts established in February— however as the first non-U.S. based program, this is a major step forward. As in the Donuts agreement, the new program will contain a number of procedural safeguards, including a requirement that clear evidence of pervasive infringement is occurring, along with a document attesting to the fact that the MPAA first attempted to resolve the situation with the name registrar directly. If, after attempting to work with its associated registrars to contact the website owner, Radix determines that the website is engaged illegal conduct it will either place the domain name on hold or else suspend it entirely.

These sorts of self-help agreements are really crucial to the future of Internet governance, and not merely for their facilitation of removing infringing content. Once ICANN becomes an independent organization that is completely untethered from the U.S. Government, it will be up to the community at large to maintain the credibility of DNS management.

And the importance of these self-help agreements is particularly acute in light of ICANN’s long standing refusal to enforce its own contractual restrictions in place with registries and registrars. As we noted in our brief:

Very likely, [ICANN’s governance structure] will be found through voluntary, private arrangements between registries, registrars, and third parties. An overarching commitment to enforcing legitimate contracts, therefore, even ones that espouse particular policy objectives, will be a core attribute of a well-organized ICANN.

In fact, far and away ICANN’s most significant failing has been the abdication of its responsibility to enforce the terms of its own contracts, particularly the Registrar Accreditation Agreement. The effect of this obstinance is that ICANN has failed to exercise its obligation to maintain a “secure, stable, [and] resilient… Internet” free of costly “pollutants” like piracy, illegal prescription drugs, and phishing sites that impose significant costs on others with relative impunity.

In March, ICANN submitted its stewardship proposal to Congress — a document that outlines how ICANN proposes to operate as an independent organization. Much criticism of the transition has focused on the possibility of authoritarian regimes co-opting the root zone file and related DNS activities. It’s in everyone’s interest to prevent clearly illegal conduct from occurring online. Otherwise, without a minimal standard of governance, the arguments for a multi-lateral government run Internet become much easier to advance.

And certainly a big part of Congress’s consideration of the transition will be whether ICANN can plausibly continue to operate as a legitimate steward of the DNS. When registries step forward and agree to maintain at least a minimal baseline of pro-social conduct, it goes a long way toward moving the transition forward and guaranteeing a free and open Internet for the future.

The ridiculousness currently emanating from ICANN and the NTIA (see these excellent posts from Milton Mueller and Eli Dourado on the issue) over .AMAZON, .PATAGONIA and other “geographic”/commercial TLDs is precisely why ICANN (and, apparently, the NTIA) is a problematic entity as a regulator.

The NTIA’s response to ICANN’s Governmental Advisory Committee’s (GAC) objection to Amazon’s application for the .AMAZON TLD (along with similar applications from other businesses for other TLDs) is particularly troubling, as Mueller notes:

In other words, the US statement basically says “we think that the GAC is going to do the wrong thing; its most likely course of action has no basis in international law and is contrary to vital policy principles the US is supposed to uphold. But who cares? We are letting everyone know that we will refuse to use the main tool we have that could either stop GAC from doing the wrong thing or provide it with an incentive to moderate its stance.”

Competition/antitrust issues don’t seem to be the focus of this latest chapter in the gTLD story, but it is instructive on this score nonetheless. As Berin Szoka and I wrote in ICLE’s comment to ICANN on gTLDS:

Among the greatest threats to this new “land rush” of innovation is the idea that ICANN should become a competition regulator, deciding whether to approve a TLD application based on its own competition analysis. But ICANN is not a regulator. It is a coordinator. ICANN should exercise its coordinating function by applying the same sort of analysis that it already does in coordinating other applications for TLDs.

* * *

Moreover, the practical difficulties in enforcing different rules for generic TLDs as opposed to brand TLDs likely render any competition pre-clearance mechanism unworkable. ICANN has already determined that .brand TLDs can and should be operated as closed domains for obvious and good reasons. But differentiating between, say .amazon the brand and .amazon the generic or .delta the brand and .delta the generic will necessarily result in arbitrary decisions and costly errors.

Of most obvious salience: implicit in the GAC’s recommendation is the notion that somehow Amazon.com is sufficiently different than .AMAZON to deny Amazon’s ownership of the latter. But as Berin and I point out:

While closed gTLDs might seem to some to limit competition, that limitation would occur only within a particular, closed TLD. But it has every potential to be outweighed by the dramatic opening of competition among gTLDs, including, importantly, competition with .com.

In short, the market for TLDs and domain name registrations do not present particular competitive risks, and there is no a priori reason for ICANN to intervene prospectively.

In other words, treating Amazon.com and .AMAZON as different products, in different relevant markets, is a mistake. No doubt Amazon.com would, even if .AMAZON were owned by Amazon, remain for the foreseeable future the more relevant site. If Latin American governments are concerned with cultural and national identity protection, they should (not that I’m recommending this) focus their objections on Amazon.com. But the reality is that Amazon.com doesn’t compromise cultural identity, and neither would Amazon’s ownership of .AMAZON. Rather, the wide availability of new TLDs opens up an enormous range of new competitive TLD and SLD constraints on existing, dominant .COM SLDs, any number of which could be effective in promoting and preserving cultural and national identities.

By the way – Amazonia.com, Amazonbasin.com and Amazonrainforest.com, presumably among many others, look to be unused and probably available for purchase. Perhaps opponents of Amazon’s ownership of .AMAZON should set their sights on those or other SLDs and avoid engaging in the sort of politicking that will ultimately ruin the Internet.