The new gTLD process seems to have leapt over yet another hurdle. An ICANN working group has released a set of policy recommendations to guide the suppression of the top level domain strings that people find “objectionable.” The Recommendations are, remarkably, pretty good. Public comment can be made here.

The ICANN “Cross-community” Working Group (CWG) was a procedurally unique combination of GAC members, GNSO representatives, and At Large Advisory Committee members. It was co-chaired by Chuck Gomes (GNSO), Cheryl Langdon-Orr (ALAC) and GAC interim Chair Heather Dryden (Canada). It was created after the Brussels ICANN meeting June 2010 in response to the US government's sudden discovery that there was no internationally recognized standard of “morality and public order” (MAPO) that could be used to censor top level domain names. The US and some other governments wanted to revise the standards to give government more flexibility to object to “sensitive” proposed top level domains. Other interests, notably the Noncommercial Stakeholders Group and the At Large, were deeply concerned about MAPO objections would be used to censor not only domain name strings, but the content associated with web sites or uses of those domains.

For example, some culturally conservative governments have expressed their opposition to the creation of a .gay top level domain. Thus, the CWG was put into the position of negotiating a new form of global content regulation. The WG recommendations are summarized briefly below.

What is interesting about the new report is that it has narrowed and tightened the standards for censoring TLD strings. The first and in some way most central recommendation is to banish the term “morality and public order” (MAPO) as the basis of objections. It will be replaced with something more grounded in international law. The group recommended that an objection, to be upheld, requires “A determination that an applied-for gTLD string would be contrary to specific principles of international law as reflected in relevant international instruments of law.” (13.1) Broader standards, such as “public interest objections,” were decisively rejected. This is good because any objection must not be arbitrary and must not be based on one national government, but rather must be rooted in established and recognized international laws. ICANN should not allow individual governments to exercise global blocking based on their local laws, norms or preferences.

Second, if individual governments have objections based on a string's alleged contradiction with specific national laws, those objections may be submitted through the Community Objections procedure using the standards outlined in version 4 of the existing applicant guidebook. The group reached full consensus that national law itself was not valid grounds for an objection.

The group also recommended important improvements in the procedural aspects of the objection process. There was consensus that ultimate resolution of the objections to a TLD should rest with the ICANN Board alone and not be delegated to a third party. The earlier proposed procedure would have outsourced the objections to a dispute resolution service of the International Chamber of Commerce(!) Why the ICC should be in charge of adjudicating free expression issues and morality and public order concerns had many of us scratching our heads. Now, panels of experts with knowledge of international law and human rights and civil liberties laws a that provide advice to the Board but do not make the decision. There was strong support for the idea that an objection, to be upheld, requires a supermajority of the board.

The CWG held an interesting debate over whether a TLD string itself could incite illegal activity. The working group improved the definition of what kind of strings would be illegal by replacing the term “incitement” with the phrase “incitement and instigation” – raising the bar to make such objections more compliant with known legal standards. Along those lines there was a fascinating debate over whether the string itself should be the only basis for an objection, or whether the nature of the applicant or other ancillary information could be used. The danger here is that ICANN could stray into content regulation or even deny people TLD strings because someone just doesn't like the applicant. There was not full consensus on how to handle this question but there was strong support for a recommendation that “The experts should conduct their analysis on the basis of the string itself. It could, if needed, use as additional context the intended purpose of the TLD as stated in the application.”

There was full consensus on a very important statement: that avoidance of national blocking of TLDs “is of little value if the result is that the opportunity to create new TLDs is unduly constrained by an objection process” and “it creates a name space that does not reflect the true diversity of ideas, cultures and views on the Internet.” This recommendation effectively repudiates the August 2010 GAC statement that governmental attempts to block controversial domains justify extensive pre-censorship of domains.

A procedural modification introduced by the group was that GAC and ALAC can collectively serve as the source of objections, and not pay any fee, but those objections must meet the same criteria and standards. The idea that GAC and ALAC-based objections should meet lower standards was rejected.

All in all, the CWG was a quick and effective multistakeholder response to the problem caused by the discredited MAPO standard. Now the Board must take these recommendations into consideration while it finalizes the new gTLD program.