Guest post: Pritz on policy vs implementation and the brother-in-law test

A current, important debate in internet governance and operation of the multi-stakeholder model asks: when do the bottom-up policy-making volunteers let go and when do the staff policy implementers take over?
Drawing that line to the satisfaction of everyone seems impossible. That is because there is no bright line — the development and implementation of policy is a task requiring the constant attention and cooperation of policy makers and implementation teams.
Is an ICANN action a policy position? Or is it a mere implementation detail? Labels almost never work, especially one-word labels. Since when are ICANN issues black or white?
We’ve stopped discussing the issues, it is easier to discuss the labels: is it “policy” or “implementation”?
The multi-stakeholder model depends on communication, consideration and collaboration
ICANN has a rich history of its stakeholders butting heads with the Board and both butting heads with the staff: long lines at microphones, speaking in hyperbole, and wringing of hands that the multi-stakeholder model has failed. ICANN also has a rich history of staff-stakeholder collaboration in the formulation of policy and in the implementation of policy.
The initial Inter-Registrar Transfer Policy was little more than a framework. Then a team of ICANN staff and TLD registry operators met over a period of months and developed the implementation model. It really didn’t matter where the policy making ended and the implementation started. There was a resolve that there should be an easy-to-follow way for registrants to transfer names and there was teamwork to get that accomplished.
There are parallels in the “real” world. In 1990, the United States enacted the American with Disabilities Act (ADA). The ADA itself was little more than a framework also: American employers should make “reasonable accommodation” for those with a “disability.” What was a reasonable accommodation? And what qualified as a disability?
Over a period of many years, those questions have been answered through a series of many court cases and regulations. But in 1990 employers were trying to figure that out.
At the time, we worked on developing clear criteria that would let our employees know what was covered by the ADA. After failing at that, we developed an approach to treat employees as you would your brother-in-law. You are deferential to your brother-in-law (because you have to face your sister) but not totally deferential (after all, he is making love to her). You just treat him better than the letter of the law requires.
So when your employee comes up to you and asks for better lighting at his workstation, you don’t try to parse whether you are required to accommodate near-sighted employees. You say, this is my brother-in-law, and I will listen to his request and carefully consider it.
This is how a public participation model works. The stakeholders are not strangers to one another and the model only works if there is mutual trust and respect.
If someone says, “I want to be heard,” she is generally listened to. (That doesn’t mean discussion is never ending. If that same person wants to be heard again, and says the same thing without change, she will be disregarded. If she continually repeats the same demand and content, she will be shunned.)
Policy versus Implementation should be Policy and Implementation
Take the example currently debated. There is a new gTLD policy element (approved in 2009) that states:

Strings must not infringe the existing legal rights of others that are recognized or enforceable under generally accepted and internationally recognized principles of law.

How does one implement that? The implementation of trademark rights protection mechanisms were developed after years of community/staff consultation, “implementability” studies, draft positions, memoranda describing potential solutions and the reasoning behind them, debate, and discussion.
When things were apparently settled, new parties joined the ICANN discussion — they were welcomed as were their opinions. ICANN was richer because there were new participants in the model. New implementation models were written. Finally, there was an indication the discussion was spent. The work was “said and done.”
Then, apparently, not all was said-and-done. After the gTLD program was launched, there were new suggestions and participants. ICANN decided to entertain those ideas. After a round of community feedback, a subset of the new suggestions was recommended by ICANN for inclusion into the implementation plan for new gTLDs.
ICANN’s policy makers, the GNSO, weighed in, agreeing with many of the conclusions but picking one of the recommendations and saying, “we’d like to talk a bit more about this one because we don’t fully understand its implications and effects.” (Unfortunately, the GNSO didn’t say exactly that, it sounded more to me like, “this item is policy and therefore it cannot be implemented without our consensus opinion.”)
Now, if my brother-in-law says he wants to talk about something some more, even if he doesn’t give a good reason, I am ready to indulge him. But ICANN did not indulge the GNSO. Now, we are in a policy versus implementation discussion: what work is the province of policy makers, and the province of implementers? The GNSO is considering ICANN Bylaw changes to ensure they are heeded.
Bylaw changes are not the lynchpin to multi-stakeholder model success. Putting rules in place for how and when to listen never work. There will always be exceptions. (Another whole piece can be written on how the rules governing communications between ICANN and its Governmental Advisory Committee have failed to facilitate the success of the multi-stakeholder model.)
Processes and procedures (as they are currently described in the Bylaws) are important. We must have clear rules for creation of consensus policy, with timelines and borders to ensure that issues are addressed and rights are respected.
But the operation of the multi-stakeholder model is more complicated than following those processes. The success of the model depends on the mutual trust and respect of the participants, and the ability to actively listen and to understand what is meant, even if that is not exactly what is said.
Rather than create new rules or discuss how one side can prevent the other from abusing its position, the volunteers, staff and Board should look inwardly to improve its own listening and communicating.
You’re my brother-in-law. I am ready (more than ready) to disagree with you, but first I am going to listen to what you have to say.
This is a guest post written by Kurt Pritz, ICANN’s former chief strategy officer. He is currently an independent consultant working with new gTLD applicants and others.

GNSO wins minor victory in Trademark+50 dispute

The ICANN board has rescheduled an important decision for trademark owners, apparently at the behest of members of the Generic Names Supporting Organization Council.
The board’s New gTLD Program Committee was due to vote June 11 on whether to approve the rejection of a Reconsideration Request filed by the Non-Commercial Stakeholders Group.
But the item has been removed from the agenda and will now instead be discussed at a new June 18 meeting that appears to have been specially scheduled for the purpose.
The rescheduling follows an appeal by GNSO Councillor Jeff Neuman directly to the committee and other senior ICANNers.
Neuman and others were concerned that a June 11 decision would preempt a discussion of the issue slated for the Council’s June 13 meeting, which would have been very bad for board-GNSO relations.
For the full background, read this post.
Essentially, Neuman and other councilors are worried that ICANN seems to be riding roughshod over the GNSO in an attempt to make a proposal known as “Trademark+50” a part of the new gTLD program.
Trademark+50 is a mechanism that will greatly expand the number of strings trademark owners can submit to the Trademark Clearinghouse and get limited protection for.
The NCSG’s Reconsideration Request had asked ICANN to reconsider its classification of the proposal as an “implementation” change that didn’t require GNSO “policy” review.
But the ICANN board’s Board Governance Committee, which adjudicates such matters, last month rejected the request in what I would describe as a sloppily argued and disconcertingly adversarial decision.
It’s now up to the New gTLD Program Committee, acting for the full board, to rubber-stamp the rejection, clearing the path for Trademark+50 to become law for new gTLD registries.
Rescheduling the decision won’t change the outcome, in my view. Trademark+50 is very probably a done deal.
But voting before the GNSO Council even had a chance to put its concerns to the board would have given fuel to the argument that ICANN ignores the GNSO when it is politically expedient to do so.
ICANN may have dodged a bullet for now, but the dispute continues.

The True Historie of Trademark+50 and the Deathe of the GNSO (Parte the Thirde)

ICANN’s decision to press ahead with the “Trademark+50” trademark protection mechanism over the objections of much of the community may not be the end of the controversy.
Some in the Generic Names Supporting Organization are even complaining that ICANN’s rejection of a recent challenge to the proposal may “fundamentally alter the multi-stakeholder model”.
Trademark+50 is the recently devised adjunct to the suite of rights protection mechanisms created specially for the new gTLD program.
It will enable trademark owners to add up to 50 strings to each record they have in the Trademark Clearinghouse, where those strings have been previously ruled abusive under UDRP.
Once in the TMCH, they will generate Trademark Claims notices for both the trademark owner and the would-be registrant of the matching domain name during the first 60 days of general availability in each new gTLD.
Guinness, for example, will be able to add “guinness-sucks” to its TMCH record for “Guinness” because it has previously won guinness-sucks.com in a UDRP decision.
If somebody then tries to register guinness-sucks.beer, they’ll get a warning that they may be about to infringe Guinness’ trademark rights. If they go ahead and register anyway, Guinness will also get an alert.
Trademark+50 was created jointly by ICANN’s Business Constituency and Intellectual Property Constituency late last year as one of a raft of measures designed to strengthen rights protection in new gTLDs.
They then managed to persuade CEO Fadi Chehade, who was at the time still pretty new and didn’t fully appreciate the history of conflict over these issues, to convene a series of invitation-only meetings in Brussels and Los Angeles to try to get other community members to agree to the proposals.
These meetings came up with the “strawman solution”, a list of proposed changes to the program’s rights protection mechanisms.
Until two weeks ago, when DI managed to get ICANN to publish a transcript and audio recording of the LA meetings, what was said during these meetings was shrouded in a certain degree of secrecy.
I don’t know why. Having listened to the 20-hour recording, I can tell you there was very little said that you wouldn’t hear during a regular on-the-record public ICANN meeting.
Everyone appeared to act in good faith, bringing new ideas and suggestions to the table in an attempt to find a solution that was acceptable to all.
The strongest resistance to the strawman came, in my view, from the very small number (only one remained by the end) of non-commercial interests who had been invited, and from the registrars.
The non-coms were worried about the “chilling effect” of expanding trademark rights, while registrars were worried that they would end up carrying the cost of supporting confused or frightened registrants.
What did emerge during the LA meeting was quite a heated discussion about whether the IPC/BC proposals should be considered merely “implementation” details or the creation of new “policy”.
That debate spilled over into 2013.
Under the very strictest definition of “policy”, it could be argued that pretty much every aspect of every new rights protection mechanism in the Applicant Guidebook is “implementation”.
The only hard policy the GNSO came up with on trademarks in new gTLDs was back in 2008. It reads:

Strings must not infringe the existing legal rights of others that are recognized or enforceable under generally accepted and internationally recognized principles of law.

Pretty much everything that has come since has been cobbled together from community discussions, ad hoc working groups, ICANN staff “synthesis” of public comments, and board action.
But many in the ICANN community — mainly registries, registrars and non-commercial interests — say that anything that appears to create new rights and/or imposes significant new burdens on the industry should be considered “policy”.
During the LA meetings, there was broad agreement that stuff like extending Trademark Claims from 60 to 90 days and instituting a mandatory 30-day notice period before each Sunrise period was “implementation”.
Those changes won’t really incur any major new costs for the industry; they merely tweak systems that already have broad, if sometimes grudging, community support.
But the attendees were split (IPC/BC on the one side, most everyone else on the other) about whether Trademark+50, among other items, was new policy or just an implementation detail.
If something is “policy” there are community processes to deal with it. If it’s implementation it can be turned over to ICANN staff and forgotten.
Because the registries and registrars have an effective veto on GNSO policy-making and tend to vote as a bloc, many others view a “policy” label as a death sentence for something they want done.
A month after the strawman meetings, in early December, ICANN staff produced a briefing paper on the strawman solution (pdf) for public comment. Describing what we’re now calling Trademark+50, the paper stated quite unambiguously (it seemed at the time):

The inclusion of strings previously found to be abusively registered in the Clearinghouse for purposes of Trademark Claims can be considered a policy matter.

Chehade had previously — before the strawman meetings — strongly suggested in a letter to members of the US Congress that Trademark+50 was not doable:

It is important to note that the Trademark Clearinghouse is intended be a repository for existing legal rights, and not an adjudicator of such rights or creator of new rights. Extending the protections offered through the Trademark Clearinghouse to any form of name (such as the mark + generic term suggested in your letter) would potentially expand rights beyond those granted under trademark law and put the Clearinghouse in the role of making determinations as to the scope of particular rights.

Personally, I doubt then-new Chehade wrote the letter (at least, not without help). It mirrors Beckstrom-era arguments and language and contrasts with a lot of what he’s said since.
But it’s a pretty clear statement from ICANN’s CEO that the expansion of Trademark Claims to Trademark+50 night expand trademark rights and, implicitly, is not some throwaway implementation detail.
Nevertheless, a day after the staff briefing paper Chehade wrote to GNSO Council chair Jonathan Robinson in early December to ask for “policy guidance” on the proposal.
Again, there was a strong suggestion that ICANN was viewing Trademark+50 as a policy issue that would probably require GNSO input.
Robinson replied at the end of February, after some very difficult GNSO Council discussions, saying “the majority of the council feels that is proposal is best addressed as a policy concern”.
The IPC disagreed with this majority view, no doubt afraid that a “policy” tag would lead to Trademark+50 being gutted by the other GNSO constituencies over the space of months or years.
But despite ICANN staff, most of the GNSO Council and apparently Chehade himself concluding that Trademark+50 was policy, staff did a U-turn in March and decided to go ahead with Trademark+50 after all.
An unsigned March 20 staff report states:

Having reviewed and balanced all feedback, this proposal appears to be a reasonable add-on to an existing service, rather than a proposed new service.
…
It is difficult to justify omission of a readily available mechanism which would strengthen the trademark protection available through the Clearinghouse. Given that the proposal relies on determinations that have already been made independently through established processes, and that the scope of protection is bounded by this, concerns about undue expansion of rights do not seem necessary.

This caught the GNSO off-guard; Trademark+50 had looked like it was going down the policy track and all of a sudden it was a pressing reality of implementation.
Outraged, the Non-Commercial Stakeholders Group, which had been the strongest (if smallest through no fault of their own) voice against the proposal during the strawman meetings filed a formal Reconsideration Request (pdf) with ICANN.
Reconsideration Requests are one of the oversight mechanisms built into ICANN’s bylaws. They’re adjudicated by ICANN’s own Board Governance Committee and never succeed.
In its request, the NCSG told a pretty similar history to the one I’ve just finished relating and asked the BGC to overturn the staff decision to treat Trademark+50 as implementation.
The NCSG notes, rightly, that just because a domain has been lost at UDRP the string itself is not necessarily inherently abusive. To win a UDRP a complainant must also demonstrate the registrant’s bad faith and lack of rights to the string at issue.
To return to the earlier example, when notorious cybersquatter John Zuccarini — an unambiguously bad guy — registered guinness-sucks.com back in 2000 he told Guinness he’d done it just to piss them off.
That doesn’t mean guinness-sucks.beer is inherently bad, however. In many jurisdictions I would be well within my rights to register the domain to host a site criticizing the filthy brown muck.
But if I try to register the name, I’m going to get a Trademark Claims notice asking me to verify that I’m not going to infringe Guinness’ legal rights and advising me to consult a lawyer.
Chilling effect? Maybe. My own view is that many people will just click through the notice as easily as they click through the Ts&Cs on any other web site or piece of software.
Either way, I won’t be able to claim in court that I’d never heard of Guinness^TM, should the company ever decide to sue me.
Anyway, the NCSG’s Reconsideration Request failed. On May 16 the BGC issued a 15-page determination (pdf) denying it.
It’s this document that’s causing consternation and death-of-the-GNSO mutterings right now.
Last week, Neustar’s lead ICANN wonk Jeff Neuman asked for the Reconsideration Request to be put on the agenda of the GNSO Council’s June 13 meeting. He wants BGC representatives to join the call too. He wrote:

This decision was clearly written by legal counsel (and probably from outside legal counsel). It was written as a legal brief in litigation would be written, and if upheld, can undermine the entire bottom-up multi-stakeholder model. If ICANN wanted to justify their decision to protect their proclamation for the 50 variations, they could have done it in a number of ways that would have been more palatable. Instead, they used this Reconsideration Process as a way to fundamentally alter the multi-stakeholder model. It not only demonstrates how meaningless the Reconsideration process is as an accountability measure, but also sends a signal of things to come if we do not step in.

He has support from other councilors.
I suspect the registries that Neuman represents on the Council are not so much concerned with Trademark+50 itself, more with the way ICANN has forced the issue through over their objections.
The registries, remember, are already nervous as hell about the possibility of ICANN taking unilateral action to amend their contracts in future, and bad decision-making practices now may set bad precedents.
But Neuman has a point about the legalistic way in which the Reconsideration Request was handled. I spotted a fair few examples in the decision of what can only be described as, frankly, lawyer bullshit.
For example, the NCSG used Chehade’s letter to Congress as an example of why Trademark+50 should be and was being considered “policy”, but the BGC deliberately misses the point in its response, stating:

The NCSG fails to explain, however, is how ICANN policy can be created through a proclamation in a letter to Congress without following ICANN policy development procedures. To be clear, ICANN cannot create policy in this fashion.

Only a lawyer could come up with this kind of pedantic misinterpretation.
The NCSG wasn’t arguing that Chehade’s letter to Congress created a new policy, it was arguing that he was explaining an existing policy. It was attempting to say “Hey, even Fadi thought this was policy.”
Strike two: the NCSG had also pointed to the aforementioned staff determination, since reversed, that Trademark+50 was a policy matter, but the BCG’s response was, again, legalistic.
It noted that staff only said Trademark+50 “can” be considered a policy matter (rather than “is”, one assumes), again ignoring the full context of the document.
In context, both the Chehade letter and the March staff document make specific reference to the fact that the Implementation Recommendation Team had decided back in 2009 that only strings that exactly match trademarks should be protected. But the BGC does not mention the IRT once in its decision.
Strike three: the BGC response discounted Chehade’s request for GNSO “policy guidance” as an “inartful phrase”. He wasn’t really saying it was a policy matter, apparently. No.
Taken as a whole, the BGC rejection of the Reconsideration Request comes across like it was written by somebody trying to justify a fait accompli, trying to make the rationale fit the decision.
In my view, Trademark+50 is quite a sensible compromise proposal with little serious downside.
I think it will help trademark owners lower their enforcement costs and the impact on registrars, registries and registrants’ rights is likely to be minimal.
But the way it’s being levered through ICANN — unnecessarily secretive discussions followed by badly explained U-turns — looks dishonest.
It doesn’t come across like ICANN is playing fair, no matter how noble its intentions.

Unrest remains despite new new gTLD contract

ICANN has proposed big changes to how it will handle premium domain names, dot-brands, mergers and acquisitions and mandatory fees in new gTLDs.
It published a new version of the proposed Registry Agreement for new gTLD operators this morning, saying that it is the product of months of “negotiations” with applicants and registries.
But some applicants and back-end providers disagree with this characterization, saying that while some registries helped ICANN with the text they have no authority to speak for all applicants.
The agreement was posted for 42 days of public comment this morning. Before it is approved by the ICANN board of directors, no new gTLD applicants will be able to sign contracts and begin to go live.
There are several major changes compared to the version in the Applicant Guidebook.
Premium domains not dead after all
In what could prove to be the most significant and controversial changes, ICANN has given registries the ability to run Founders Programs and premium name schemes without interference from trademark owners.
New text in the contract will let them self-register up to 100 names “necessary for the operation or the promotion of the TLD” and release those names to third parties if they want.
This appears to be a way around the fear that mandatory Sunrise periods could thwart registries’ plans to sign up anchor tenants to the gTLDs, a crucial launch marketing tactic for many.
The new RA also appears to give broad powers to the registry to allocate premium domain names at will.

Registry Operator may withhold from registration or allocate to Registry Operator names (including their IDN variants, where applicable) at All Levels in accordance with Section 2.6 of the Agreement. Such names may not be activated in the DNS, but may be released for registration to another person or entity at Registry Operator’s discretion.

There does not appear to be a numerical limit on how many domains can be reserved in this way.
Hypothetically, this might allow a registry to reserve the entire dictionary (or dictionaries) at launch, preventing holders of trademarks on generic terms grabbing the matching names during Sunrise.
The still-draft Trademark Clearinghouse rules will also play a part here, but from the RA it looks like registries have just been handed a massively flexible reservation tool.
If my initial interpretation is correct, I expect the trademark lobby will have strong view here.
Concessions for dot-brands
New text in the agreement makes it clearer that ICANN has no plans to redelegate dot-brand gTLDs to third parties after the Registry Agreement expires or is terminated.
This means, for example, that if L’Oreal decides to stop using .loreal at some point in future, ICANN very probably won’t give .loreal to a competitor. The new text is:

(i) ICANN will take into consideration any intellectual property rights of Registry Operator (as communicated to ICANN by Registry Operator) in determining whether to transition operation of the TLD to a successor registry operator

It’s probably not rigid enough language to satisfy some lawyers’ wishes, but I think it does enough to convey the spirit of ICANN’s intentions.
ICANN is of course mainly concerned that dead gTLDs don’t leave registrants with dead domain names, but if there are no registrants I can’t imagine why it would want to redelegate.
Lower fees for registries
Newly added text in the RA specifies that registries must pay ICANN a $5,000 one-off fee (per TLD) to use the new Trademark Clearinghouse, plus with $0.25 per domain that uses its services.
Domains registered under Sunrise periods or which trigger Trademark Claims alerts would incur this one-time fee, which appears to have been reduced from the $0.30 previously discussed.
These fees will actually be passed on to the Trademark Clearinghouse operators (Deloitte and IBM), for which ICANN has agreed to manage billing in order to keep costs down.
In addition, the RA now clarifies that the registry operator’s regular fixed fees to ICANN of $6,250 a quarter only kick in from the date that the gTLD hits the DNS root, not the date of contract signing. That could save registries up to a year’s worth of fees, if they’re late to delegation.
M&A approvals
There are also changes to the way ICANN plans to approve of mergers and acquisitions among registries.
First, it will be much easier for the contract to be passed around within a corporate holding group. The RA now states:

Registry Operator may assign this Agreement without the consent of ICANN directly to a wholly-owned subsidiary of Registry Operator, or, if Registry Operator is a wholly-owned subsidiary, to its direct parent or to another wholly-owned subsidiary of its direct parent, upon such subsidiary’s or parent’s, as applicable, express assumption of the terms and conditions of this Agreement

This change would seem to enable portfolio applicants that have applied for many gTLDs each under separate shell company names (Donuts, for example) to consolidate their contracts under a single parent.
What I don’t think it does is allow for contention set resolution based on joint ventures (which are obviously not “wholly owned”), such as what Uniregistry and Top Level Domain Holdings announced they had agreed to yesterday.
The new RA also states that ICANN must approve subcontracting deals the registry inks for any of the five “critical functions” (EPP, DNS, DNSSEC, Whois and escrow).
Unilateral amendments are gone
The controversial “unilateral right to amend” that ICANN wanted to grant itself — essentially an emergency power to change the contract almost at whim and over the objections of registries — is gone.
It’s been replaced with a convoluted series of procures almost identical to those found in the proposed final version of the 2013 Registrar Accreditation Agreement currently open for comment.
Registries would get the ability to punt the changes to a GNSO Policy Development Process, submit alternative amendments, take ICANN to arbitration or request exemptions, under the new rules.
While the new provisions still give ICANN the ability to force through unpopular changes under certain circumstances, a lot more engagement by registries is envisaged so “unilateral” is probably not a good word to use any more.
So is the deal final or not?
ICANN said in a blog post: “The proposed agreement is the result of several months of negotiations, formal community feedback, and meetings with various stakeholders and communities.”
It added:

We have come a long way since February 2013 when we posted a proposed Revised New gTLD Registry Agreement for public comment. A new and highly spirited sense of mutual trust has catapulted us into a fresh atmosphere of collaboration, which in turn has led to a consistently more productive environment. The spirit of teamwork, productive dialogue and partnership that has underpinned this negotiation process is tremendously heartwarming, as it has allowed us to bring to fruition a robust contractual framework for the New gTLD Program.

But some are worried that ICANN seems to be portraying the RA as equivalent to the Registrar Accreditation Agreement, which was subject to 18 months of talks with a negotiating team representing registrars.
The registries’ Registry Agreement Negotiating Team (RA-NT), on the other hand, was formed less than three weeks ago during ICANN’s meeting in Beijing, and did not have the authority to speak for all applicants.
The RA-NT said in a statement published by ICANN:

The RA-NT agreed to review the new gTLD Registry Agreement with ICANN staff in an effort to minimize some of the more controversial aspects of the Agreement for applicants as a whole. While participants reflected a variety of perspectives, the team did not “represent” or have any authority to “speak for” new gTLD applicants generally, or any group of applicants.

ARI Registry Services CEO Adrian Kinderis told DI:

My fears (and frustrations) come from the fact that ICANN staff have made it sound like they have reached the same point in the process. “It is done”. It most certainly isn’t “done”. They need to understand that the negotiation is actually still very much active and all of the community should feel like their opinions and feedback will be considered in the development of the “final draft”.

The draft RA is now open for public comment until June 11.
That would give ICANN about a month to synthesize all the comments, make any changes, and put the deal to its board of directors for approval during the meeting in Durban, South Africa, this July.

Will the Trademark Clearinghouse kill off premium domains?

Rules proposed for the new Trademark Clearinghouse threaten to cut off some of new gTLD registries major sources of early revenue, according to registry providers.
Premium domain sales and founders programs are among the now industry-standard practices that would be essentially banned under the current draft of the TMCH rules, they say.
The potential problems emerged in a draft TMCH Requirements document circulated to registries 10 days ago and vigorously discussed during a session at the ICANN meeting in Beijing last week.
The document lists all of the things that new gTLD registries must and must not, and may and may not, do during the mandatory Sunrise and Trademark Claims rights protection launch periods.
One of the bits that has left registries confused is this:

2.2.4 Registry Operator MUST NOT allow a domain name to be reserved or registered to a registrant who is not a Sunrise-Eligible Rights Holder prior to the conclusion of the Sunrise Period.

What this means is that trademark owners get first dibs on pretty much every possible string in every gTLD.
“Trademark owners trump everything,” Neustar business affairs veep Jeff Neuman said during the Beijing meeting. “Trademark owners trump every possible use of every possible name.”
It would mean, for example, that if a new gTLD wanted to allocate some names to high-profile anchor tenants during a “founders program”, it would not be able to do so until after the Sunrise was over.
Let’s say the successful applicant for .shop wants to reserve the names of hundreds of shop types (book.shop, food.shop, etc) as premium names, to allocate during its founders program or auction later.
Because the .shop Sunrise would have to happen first, the companies that the own rights to, for example, “wallpaper” or “butcher” (both real US trademarks) would have first rights to wallpaper.shop and butcher.shop, even if they only planned to defensively park the domains.
Because there’s likely to be some degree of gaming (there’s a proof-of-use requirement, but the passing threshold is pretty low), registries’ premium lists could be decimated during Sunrise periods.
If ICANN keeps its TMCH Requirements as they are currently written, new gTLD registries stand to lose a lot of early revenue, not to mention control over launch marketing initiatives.
However, if ICANN were to remove this rule, it might give unscrupulous registries the ability to circumvent the mandatory Sunrise period entirely by placing millions of strings on their premium lists.
“Registries should have discretion to schedule their start-up phases according to their business plans so long as rights protection processes are honored, so that’s the balancing we’ve tried to do,” ICANN operations & policy research director Karen Lenz said during Beijing.
“It’s trying to allow registries to create requirements that suit their purposes, without being able to hollow out the rights protection intention,” she said.
The requirements document is still just a draft, and discussions are ongoing, she added.
“It’s certainly not our intention to restrict business models,” Lenz said.
Registries will get some flexibility to restrict Sunrise to certain registrants. For example, they’ll be able to disqualify those without an affiliation to the industry to which the gTLD is targeted.
What they won’t be able to do is create arbitrary rules unrelated to the purpose of the TLD, or apply one set of rules during Sunrise and another during the first 90 days of general availability.
The standard Registry Agreement that ICANN expects all new gTLDs to sign up to does enable registries to reserve or block as many names as they want, but only if those names are not registered or used.
It seemed to be designed to do things like blocing ‘sensitive’ strings, rather like when ICM Registry reserved thousands of names of celebrities and cultural terms in .xxx.
The Requirements document, on the other hand, seems to allow these names being released at a later date. If they were released, the document states, they’d have to be subject to Trademark Claims notices, but not Sunrise rules.
While that may be a workaround to the premium domains problem, it doesn’t appear to help registries that want to get founders programs done before general availability.
It seems that there are still many outstanding issues surrounding the Trademark Clearinghouse — many more than discussed in this post — that will need to be settled before new gTLDs are going to feel comfortable launching.

Trademark Clearinghouse to get tested out on three existing TLDs

Three already-live TLDs are going to use the Trademark Clearinghouse to handle sunrise periods, possibly before the first new gTLDs launch.
BRS Media is set to use the TMCH, albeit indirectly, in its launch of third-level domains under .radio.am and .radio.fm, which it plans to launch soon as a budget alternative to .am and .fm.
The company has hired TM.Biz, the trademark validation firm affiliated with EnCirca, to handle its sunrise, and TM.biz says it will allow brand owners to leverage Clearinghouse records.
Trademark owners will be able to submit raw trademarks for validation as in previous sunrises, but TM.Biz will also allow them to submit Signed Mark Data (SMD) files, if they have them, instead.
Encrypted SMD files are created by the TMCH after validation, so the trademarks and the strings they represent are pre-validated.
There’ll presumably be some cost benefit of using SMD files, but pricing has not yet been disclosed.
Separately, Employ Media said today that it’s getting ready to enter the final stage of its .jobs liberalization, opening up the gTLD to essentially any string and essentially any registrant.
The company will also use the TMCH for its sunrise period, according to an ICANN press release, though the full details and timing have not yet been announced.
Unusually, .jobs is a gTLD that hasn’t already had a sunrise — its original business model only allowed vetted company-name registrations.
The TMCH is already accepting submissions from trademark owners, but it’s not yet integrated with registries and registrars.

Six big reasons we won’t see any new gTLD launches until Q3

ICANN’s announcement of a big media bash in New York on April 23, to announce the launch of new gTLDs, has gotten many people thinking the first launches are imminent.
Wrong.
We’re not going to see any new gTLD domains on sale until the third quarter at the earliest, in my view, and here are a few good reasons why.
April 23 is just a PR thing
ICANN has said that April 23 is primarily about awareness-raising.
Not only does it hope to garner plenty of column inches talking about new gTLDs — helping the marketing efforts of their registries — it also hopes to ceremonially sign the first Registry Agreements.
I think CEO Fadi Chehade’s push to make the industry look more respectable will also play a part, with the promotion of the Registrant Rights and Responsibilities document.
But there’s never been any suggestion that any strings will be delegated at that time, much less go live.
The contracts are still hugely controversial
If ICANN wants to sign a Registry Agreement on April 23, it’s going to need a Registry Agreement to sign.
Right now, applicants are up in arms about ICANN’s demand for greater powers to amend the contract in future.
While ICANN has toned down its proposals, they may still be unacceptable to many registries and gTLD applicants.
Applicants have some impetus to reach agreement quickly — because they want to launch and start making money as soon as possible.
But ICANN wants the same powers added to the 2013 Registrar Accreditation Agreement, and registrars are generally less worried about the speedy approval of new gTLDs.
ICANN has tied the approval of the RA and the RAA together — only registrars on the new RAA will be able to sell domains in new gTLDs.
Chehade has also made it clear that agreement on the new RAA is a gating issue for new gTLD launches.
If registries, registrars and ICANN can’t settle these issues in Beijing, it’s hard to see how any contracts could be signed April 23. The first launch would be delayed accordingly.
GAC Advice might not be what we’re expecting
GAC Advice on New gTLDs is, in my view, the biggest gating issue applicants are facing right now.
GAC Advice is an integral part of the approval process outlined in the Applicant Guidebook and ICANN has said many times that it cannot and will not sign any contracts until the GAC has spoken.
But what does that mean from a process and timing point of view?
According to the Applicant Guidebook, if an application receives GAC Advice, it gets shunted from the main evaluation track to the ICANN board of directors for consideration.
It’s the only time the ICANN board has to get directly involved with the approval process, according to the Guidebook’s rather complex flow-charts.
GAC Advice is not an automatic death sentence, but any application the GAC is unanimously opposed to stands a very slim chance of getting approved by the board.
Given that ICANN is has said it will not sign contracts until it has received GAC Advice, and given that it has said it wants to sign the first contract April 23, it’s clearly expecting to know which applications are problematic and which are not during the next three weeks.
But I don’t think that’s necessarily going to happen. The GAC moves slowly and it has a track record of missing ICANN-imposed deadlines, which it often seems to regard as irksome.
Neither ICANN nor the GAC have ever said GAC Advice on New gTLDs will be issued during next week’s public meeting in Beijing. If a time is given it’s usually “after” or “following” Beijing.
And I don’t think the GAC, which decided against holding an inter-sessional meeting between Toronto and Beijing, is remotely close to providing a full list of specific applications of concern.
I do think a small number of slam-dunk bad applications – such as DotConnectAfrica’s .africa bid – will get Advised against during or after the Beijing meeting.
But I also think the GAC is likely to issue Advice that is much broader, and which may not provide the detail ICANN needs to carry the process forward for many applicants.
The GAC, in its most recent (delayed) update, is still talking about “categories” of concern – such as “consumer protection” and “geographical names” – some of which are very broad indeed.
Given the limited amount of time available to it in Beijing, I think it’s quite likely that the GAC is going to produce advice about categories as well as about individual applications.
And, crucially, I don’t think it’s necessarily going to give ICANN a comprehensive list of which specific applications fall into which categories.
If the GAC decides to issue Advice under the banner of “consumer protection”, for example, somebody is going to have to decide which applications are captured by that advice.
Is that just strings that relate to regulated industries such as pharmaceuticals or banking? Or is it any string that relates to selling stuff? What about .shop and .car? Shops and cars are “regulated” by consumer protection and safety laws in most countries.
Deciding which Advice covered which applications would not be an easy task, nor would it be a quick one. I don’t think the GAC has done this work yet, nor do I think it will in Beijing.
For the GAC to reach consensus advice against specific applications will in some cases require GAC representatives to return to their capitals for guidance, which would add delay.
There is, in my view, a very real possibility of more discussions being needed following Beijing, just in order to make sense of what the GAC comes up with.
The new gTLD approval process needs the GAC to provide a list of specific applications or strings with which it has concerns, and we may not see that before April 23.
ICANN may get a short list of applications that definitely do have Advice by then, but it won’t necessarily know which applications do not, which may complicate the contract-signing process.
The Trademark Clearinghouse still needs testing
The Trademark Clearinghouse is already, in one sense, open for business. Trademark owners have been able to submit their marks for validation for a couple of weeks now.
But the hard integration work has not been done yet, because the technical specifications the registries and registrars need to interface with IBM’s TMCH database have not all been finalized.
When the specs are done (it seems likely this will happen in the next few weeks), registries and registrars will need to finish writing their software and start production testing.
ICANN’s working timetable has the TMCH going live July 1, but companies that know much more than me about the technical issues at play here say it’s unlikely that they’ll be ready to go live with Sunrise and Trademark Claims services before August.
It’s in everyone’s interests to get all the bugs ironed out before launch.
For new gTLD registries, a failure of the centralized TMCH database could mean embarrassing bugs and downtime during their critical launch periods.
Trademark owners and domain registrants may also be concerned about the potential for loopholes.
For example, it’s still not clear to some how Trademark Claims – which notifies registrants when there’s a clash between a trademark and a domain they want – will interact with landrush periods.
Does the registrant only get a warning when they apply for the domain, which could be some weeks before a landrush auction? If so, what happens if a mark is submitted to the TMCH between the application and the auction and ultimate registration?
Is that a loophole to bypass Trademark Claims? Could a registrant get hit by a Claim after they’ve just spent thousands to register a domain?
These are the kinds of things that will need to be ironed out before the TMCH goes fully live.
There’s a sunrise notice period
The sunrise period is the first stage of launch in which customers get to register domain names.
Lest we forget, ICANN recently decided to implement a mandatory 30-day notice period for every new gTLD sunrise period. This adds a month to every registry’s go-live runway.
Because gTLD sunrise periods from now on all have to use the TMCH, registries may have to wait until the Clearinghouse is operational before announcing their sunrise dates.
If the TMCH goes live in July, this would push the first launch dates out until August.
Super-eager registries may of course announce their sunrise period as soon as they are able, and then delay it as necessary to accommodate the TMCH, but this might carry public relations risks.
Verisign’s security scare
It’s still not clear how Verisign’s warning about the security risks of launching new gTLDs on the current timetable will be received in Beijing.
If the GAC reckons Verisign’s “concerns” are valid, particularly on the issue of root zone stability, ICANN will have to do a lot of reassuring to avoid being advised to delay its schedule.
Could ICANN offer to finish off its work of root zone automation, for example, before delegating new gTLDs? To do so would add months to the roll-out timetable.

First seven TMCH agents approved

The Trademark Clearinghouse has listed its first seven approved trademark submission agents.
These are the companies that are able to secure bulk discounts for submitted marks to the TMCH.
The first seven are NetNames, Corporation Service Company, SafeNames, MailClub, Key-Systems, Net-Chinese, which are all domain registrars too, and the Law Offices of S.J. Christine Yang, which isn’t.
Holders of large trademark portfolios are also allowed to become agents in their own right, but most TMCH submissions are expected to be carried out via these third parties.
The full list is here.
The cost of submitting trademarks, if you’re an agent, can be worked out using the TMCH Cost Calculator.

Loophole gives trademark owners unlimited Clearinghouse records

Trademark owners will be able to add potentially thousands of strings to the Trademark Clearinghouse due to a recently introduced loophole, it emerged last night.
ICANN recently said that it will allow mark holders to add up to 50 strings related to their trademarks to their TMCH records, if the strings have been abused in the past.
It was one of the controversial “strawman” proposals that ICANN decided to adopt earlier this month.
Companies would be able to get protection for “mark+keyword” strings, for example, if a UDRP decision or court ruling had previously found that the strings had been cybersquatted.
The 50-string cap appeared to have been picked rather arbitrarily, but it turns out it’s more-or-less irrelevant anyway.
ICANN confirmed on its webinar for new gTLD applicants last night that the limit is 50 additional strings per entry in the Clearinghouse, not 50 strings per trademarked string.
What this means is that a company that has registered its trademark in multiple jurisdictions will be able to get 50 extra strings for each of those marks it enters into the Clearinghouse.
If Apple had a registered mark for “Apple” in the US and a registered mark for “Apple” in Bolivia, it would be able to submit both to the Clearinghouse and get an additional 100 “apple+keyword” records.
If it had the mark registered in 100 countries, it could put up to 5,000 more strings in the Clearinghouse.
Each string could be used to generate Trademark Claims notices, but not to secure registrations during Sunrise periods.
The apparent loophole and its implications were raised by Reg Levy of Minds + Machines during last night’s ICANN call.
In practice, the number of additional strings mark holders would qualify for would be capped by the number of trademark jurisdictions in the world and/or the number of UDRP decisions they’d won.
Few companies have secured more than a few hundred domains at UDRP to date, meaning it won’t be too difficult for trademark owners to get Trademark Claims protection for basically any previously cybersquatted string.

ICANN to pay $2 million to keep Trademark Clearinghouse “free” for registrars

ICANN is putting its money where its mouth is when it comes to helping new gTLDs be successful, committing $2 million to keep Trademark Clearinghouse access “free” for registrars.
While TMCH pricing for trademark owners is now well-publicized, ICANN COO Akram Attalah last night revealed some of the fees for new gTLD registries and registrars.
Registries will have to pay a one-time fee of $5,000 per TLD to access the TMCH, he said.
That was reduced from $10,000 during talks with TMCH back-end provider IBM after ICANN promised to handle billing and administration, he said.
There’s also going to be a $0.30 fee for each domain that matches a TMCH record registered during Sunrise and Trademark Claims periods, he added. The specifics on this fee were a little fuzzy.
But registrars won’t have to pay a penny, it seems. Attalah said that ICANN will pay IBM $2 million to make sure the Clearinghouse is accessible and free for registrars.
“ICANN will pay $400,000 per year for five years to keep the TMCH up and running and that provides free access to all registrars,” he said on last night’s new gTLDs update webinar.
It won’t be completely free for registrars, of course.
Registrars will have to do some implementation work to support the new Trademark Claims and Sunrise specs, but the absence of fees gives them one less excuse to avoid the two rights protection periods.