Registration details are here.

In the comment’s below the article, Alex Bligh raises some relevant technical objections. However this proposal also raises some interesting policy questions:

• If ISPs are to be required to provide such a service just so third parties can sell telephony services, shouldn’t the VoIP providers be required to reimburse the ISPs? If so, how could that be accomplished, especially considering some VoIP providers are based overseas?
• If ISPs ran a location service, to what other uses might it be put? Could this become a new revenue stream? Might the ability to obtain accurate location information about a user be employed to enable segmentation of the Internet, in the same way other markets are geographically segmented? For example, should non-Americans be allowed to access NBC’s online coverage of the Olympics, when European broadcasters have paid for exclusive coverage in their jurisdiction?
• In this context, what control should a user be given about their location? Should they be able to make their ISP lie about their location? Should they be able to force their ISP to refuse to disclose their location, or should the ISP be able to obtain a notional consent from the user through the Terms of Service? If so, should location services be opt-in or opt-out for the user?

However in the first instance the questions are likely to relate to protecting users seeking blue-light services. It’s entirely possible the wider ramifications will only be considered when it’s too late to do much about it.

The decision endorses a 2007 report from GNSO Council, an ICANN structure that makes recommendations to the ICANN Board on gTLD policy.

Less attention has been given to one of the new tests ICANN will use when considering whether to approve a new gTLD, contained in GNSO’s sixth recommendation:

Strings [meaning, new top level domain names] must not be contrary to generally accepted legal norms relating to morality and public order that are recognized under international principles of law.

The report goes on to amplify on what it means by “generally accepted legal norms relating to morality and public order”:

Examples of such principles of law include, but are not limited to, the Universal Declaration of Human Rights (UDHR), the International Covenant on Civil and Political Rights (ICCPR), the Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW) and the International Convention on the Elimination of All Forms of Racial Discrimination, intellectual property treaties administered by the World Intellectual Property Organisation (WIPO) and the WTO Agreement on Trade-Related Aspects of Intellectual Property (TRIPS).

Quite why intellectual property is included as an issue of “morality and public order” alongside the Universal Declaration of Human Rights isn’t explained, and probably owes more to the lobbying power of the American music and film industry associations than anything else. That aside, not everbody is comfortable with ICANN making decisions on “morality and public order”.

ICANN Board member Wendy Seltzer speaking for the At-Large Community (ALAC), that represents ordinary end users, commented:

[ALAC] expressed concern that putting these criteria into the gTLD approval process, even as opportunities for objection, injects ICANN into the business of making morality and public order decisions, or injects that into ICANN’s processes in a way that, as ALAC put it, debases the ICANN process. And at-large does not want to see ICANN put into the business of adjudicating or even delegating the adjudication of morality or public order or community support. And so we hope that in implementation, these criteria can be kept sufficiently narrow so that they are both administrable and understandable and so that they do not involve ICANN, the organization, in making, or allowing to be made, determinations about any claim to generally accepted morality principles.

ICANN Board member Professor Susan Crawford agreed, going on to say:

[N]either national governments acting as sovereigns nor intergovernmental organizations acting as representatives of governments should participate in management of Internet names and addresses.
[…]
This wasn’t done out of enthusiasm for the free market alone. The idea was also to avoid having sovereigns use the Domain Name System for their own content, control, desires. To avoid having the Domain Name System used as a choke point for content. Recommendation 6, which is the morality and public order recommendation, represents quite a sea change in this approach, because the recommendation is that strings must not be contrary to generally acceptable legal norms relating to morality and public order that are recognized under international principles of law. That’s the language of the recommendation.
Now, if this is broadly implemented, this recommendation would allow for any government to effectively veto a string that made it uncomfortable. Having a government veto strings is not allowing the private sector to lead. It’s allowing sovereigns to censor.

In the formal discussions, these issues are mainly debated in the abstract, but two key examples are bandied about in private: .jihad (which even the anti-censorship USA seems keen to prohibit) and .nazi (which is an example dear to the hearts of some European governments with strict anti-Nazi laws).

Civil libertarians supporting Susan Crawford’s line argue that if governments are able to pressure ICANN into prohibiting .jihad (which has perfectly non-violent meanings in Islam as well as the terrorist connotations it has recently acquired in the West), then can a prohibition on .falun-gong be far behind?

Traditionalists among the Internet technical community might be less impressed with the cry to protect freedom of expression in top level domains, arguing that domain names - let alone top level domain names - are intended as identifiers in an addressing scheme, not as a medium of expression at all. However, even on this analysis there is cause for concern about the “morality string criterion”

• Is it possible for a short phrase such as would be valid as a top level domain name to constitute an incitement to violence, or other generally accepted breach of public order, in and of itself? Does it not depend on how the domain is used? Does the objection to .nazi lie not so much in its identification of content that might relate to Nazi ideology and Nazis in history, but that it might be used by people sympathetic to the ideology?
• Whether or not it is possible for a domain to inherently infringe principles of morality and public order, doesn’t such a rule invite ICANN to investigate how such a domain might be used? Is such an invitation a good idea? Do we really think ICANN is well equipped to perform this role?
• If we accept that ICANN should consider the likely use of a top level domain, and weigh that against principles of morality and public order - not to mention intellectual property law - before deciding whether such a domain should exist, why should it stop there? Why shouldn’t ICANN require the registries of gTLDs (including .com) to do the same at the second level? ICANN can impose terms on such registries by contract; the only thing that restrains it is a view that this is not ICANN’s proper role or purpose. If we accept the principle that ICANN can adjudicate globally “generally accepted legal norms relating to morality and public order”, why not require gTLD registries to enforce these principles at the second level? And why stop with new domains: wouldn’t actual proof of “infringing” use be even more damning than speculation about how a new domain might be used in the future?

The string criteria debate may attract less attention than the creation of new TLDs, and may not immediately affects as many people as the introduction of Internationalised Domain Names. Nonetheless, history may yet come to view this as a watershed moment when the world first acquired a global Internet content regulator.

BT’s apparent action is in line with similar moves from Virgin Media, but contrary to the position taken by TalkTalk, which believes it should not take sides against its own customers on the basis of a complaint raised by a third party.

BT has declined to comment on The Register’s story.

In the section, “Better Regulation”, the government discloses a hardening of attitudes towards industry self-regulation, despite Byron’s support for a self-regulatory approach. The government states that

• “[W]here companies wish to take credit for having signed up to self-regulatory standards, they should also sign up to processes against which their performance can be independently assessed. All codes of practice will need to set out a process for monitoring signatories’ performance including when expectations are not being met”
• “Failing to comply with a code should be damaging to a company’s reputation and provide a strong incentive for industry to implement safety measures.”
• “The Byron Review is clear that, as with any other self-regulatory arrangement, members of the public should be able to complain if they believe a code is not being implemented. The relevant UKCCIS working groups will need to identify an independent person or body who is competent to hear these complaints”

“Given the serious privacy concerns raised by the sophisticated ad-serving technology Charter Communications planned to test market, I am pleased to hear that the company has decided to delay implementation of this program, which electronically profiled individual consumer web usage. I urge other broadband companies considering similar user profiling programs to similarly hold off on implementation while these important privacy concerns can be addressed”

The scheme, like schemes being promoted in the UK by Phorm, would entail analysing users’ web browsing “clickstreams” to create the user profiles used to select adverts.

A few points do stand out as particularly notable:

• In a document that mainly considers policy objectives only in the broadest possible terms, two very specific and technical policies are endorsed:
• a call for governments and large private sector users to switch to IPv6; and
• a call to “facilitate the introduction of IDNs [internationalised domain names] while ensuring the integrity and stability of the Internet”. IDNs are domain names tht support alternative character sets, enabling (for example) Chinese, Arabic and Danish language domain names to be displayed in their normal characters. This idea is widely supported in principle, but has met practical resistance on the grounds that technical issues might make IDNs problematic for security and infrastructure resilience - hence the qualification in the above call. For example, phishing might become even more dangerous if visually similar versions of well-known domain names resolved to different locations because they’re in a different character set.
• The Ministers declare that the Internet Economy “will strengthen our capacity to improve the quality of life for all our citizens” not only through a range of economic outcomes, but also by “enabling new forms of civic engagement and participation that promote diversity of opinions and encourage transparency, accountability, privacy and trust”. This, together with a declaration that refers to “creating opportunities for new economic and social activities”, is notable as good Internet policy is often considered a “buiness” or “economic” issue, to be balanced against social policy interests that are often portrayed as being in tension with the Internet rather than in harmony with it.
• In a paragraph begining “We agree that our challenges are …”, the second challenge listed is to “Promote Internet-based innovation, competition, and user choice”. Later on Ministers commit to policies that “uphold the open, decentalised and dynameic nature of the Internet and the development of technical standards that enable its ongoing expansion and contribute to innovation, interoperability, participation and ease of access”. All this is very welcome to those who, like LINX, support these objectives, and feel that ISPs are too often beleagured by those who want ISPs to act as “gatekeepers to the Internet” or “Internet police”, a role that would seriously undermine the Internet’s “dynamic nature” as well as innovation, interoperability, participation and ease of access.
• The pressure to address intellectual property issues is never far from these meetings, and this declaration too includes a call for policies and laws that “Ensure respect for intellectual property rights”. However there is nothing in this statement that places that interest above the other important interests it recognises, nor does it suggest unhelpful policies in support of that objective. Instead, it argues that “efforts to combat digital piracy” should be combined “with innovative approach which provide creators and rights holders with incentives to create and disseminate works in a manner that is beneficial to creators, users, and our economies as a whole” - language which appears remarkably sympathetic to the view that a large part of the problem of online copyright infringement lies with the rightsholders own licensing policies.
• The call to “Encourage collaboration between governments, the private sector, civil society and the Internet technical community to build an understanding of the impact of the Internet on minors…” endorses a four-stakeholder category model. This has previous been discussed at the Internet Governance Forum and elsewhere, although the IGF has so far only formally recognised three stakeholder classes, omitting the Internet technical community.

As an assessment, documents such as OECD Ministerial declarations may lack the excitement of more widely reported and contentious policy demands, but they are quietly important as an antidote to the excesses of headline-driven policy. The Seoul Declaration contains little to criticise, and much to like.

Update: See also comments from Geoff Huston.