Open Standards International Symposium, Yale Law School
February 3rd, 2007

The most peculiar social difference shown by the Internet, when compared to previous telecommunication systems, is the distributed, unregulated, decentralized allocation of the power to create, both in respect to content and to technology. Most of the technologies we use today over the Internet, ranging from the World Wide Web to chats and peer-to-peer download of media files, were not created by network operators, by governmental agencies, or by the investments of ICT corporations; they were rather invented and initially developed by individuals, often as a side project to their actual job, or just as “something fun”.

This shift in models has a deep impact on the role of technical standardization as well – and not just because more and more standards (starting again from HTML and HTTP) are initially created in a somewhat informal manner by one or a few individuals, and only later become formalized both in terms of theorization and of development policies.

A significant percentage of the billion of individuals who use the Internet – between 40 and 60 per cent, according to the latest data by Creative Commons – creates content or, less frequently, technology. Consequently, the refinement and adoption of new standards, which used to be a matter for a few specialized technologists meeting in agencies like the ITU or in corporate working groups, becomes a live activity, happening every day in many corners of the Internet. Technical standards, which used to be embedded into technological objects and invisible to users, now become the visible subject and instrument of the everyday activity of millions of people. The age of the mass technical standard, when even a child can spell 'h-t-t-p', and when a 20-year-old hacker from India can join IETF groups with Microsoft and IBM engineers, has finally come.

At the same time, in a global environment with no centralized planning of new developments, technical standards gain the factual ability to become law in themselves. The supposedly technical decision by an industry-only consortium to add a “Region” field to the DVD format has a huge impact on the global media market and on the flow of culture. The supposedly technical decision on whether the standards for internationalized domain names are mature enough to be incorporated into the root zone file of the DNS tremendously affects the penetration of the Internet in certain parts of the world. Technical standards now have a significant political value; the border between standards and policies blurs.

As the awareness of this phenomenon grows, it appears more and more compelling that the purpose and nature of technical standardization processes is collectively revised. Standards do not just have to be technically effective; they need to be democratic. Actually, important standards were revised or integrated in the last decade, trading off some effectiveness in exchange for better support for the public interest: as an example, consider the change that was introduced in the original architecture of the GSM mobile telephony platform to allow for number portability, which helped in opening up the related market and eating into the pre-existing monopolies.

The key issue is then how to ensure that standards take into account the global public interest: the interest of the Internet and of the global society of planet Earth as a whole, as opposed to that of specific countries, companies and social groups. Governments are bound to the specific interest of their country; companies are focused on maximizing their profits. The only group that can represent the global public interest of the Internet is the aggregation of the users.

Thus, direct participation by individual users in the definition of new standards is vital to prevent technical norms from being used to disrupt the original architecture of the Internet, removing the users' freedom to innovate, and to ensure that the global public interest is kept into account.

One first necessary action item is to ensure that all closed standardization processes, be them intergovernmental or industry-run, are opened up to user participation, as well as participation by all other stakeholders who wish to do so, including developing countries. This point was already raised by the United Nations' Working Group on Internet Governance, and approved as a principle¹ at the U.N. World Summit on Information Society, which even urged international organizations to put it in practice².

However, a practical problem of representation arises. Governments and companies can afford direct participation, and anyway have reasonably clear paths to their representation. Users, instead, usually hold a stake that is too small for each of them to act individually, though it becomes huge if we consider users as an aggregated entity. The Internet, by lowering practical barriers for participation and self-organization, has fostered an impressive growth of the figure of the “power user”, the individual who cares enough to act in first person; still, can any number of “power users” showing up at a meeting or in a working group be considered representative? How can capture or the so-called “astroturf” participation of fake users be avoided?

While there is no simple and firm reply to this question, the experience at ICANN – the first multi-stakeholder technical administration entity of the Internet – has brought forward several considerations. First of all, user participation will always be in small numbers, will vary over time depending on media coverage and on the specific issues, and will never be particularly timely or predictable; these are not good reasons to diminish its value. In any case, even if opening doors – allowing individuals to join the discussions, waiving fees, publishing information – is a prerequisite, it is not sufficient that doors are open to claim that participation requirements have been met; there is the need to nurture it with appropriate practices, online tools, and subsidizing physical presence by users and civil society advocates when necessary.

Nonetheless, effective user participation is not just a matter of principle; it is necessary in practice. Reality shows that, exactly as for policies, when technologies that do not meet the requirements and viewpoints of the users are deployed, they are doomed to fail, or to cause more trouble than they solve. Think, for example, of all the technologies for prescriptive digital rights management that have been tried in the last years, and that ultimately failed. In the end, from the industry's point of view, the sheer cost of trying to deploy technologies and policies against the opposition of the users is possibly higher than the cost of compromising with user interests in the first place. Money spent bringing users to the discussion table is money well spent; companies do this ordinarily in their marketing activities, it just makes sense to do the same with standards.

User participation is not about representation, but about advocacy: about putting good ideas on the table, and providing early feedback in the technical design cycle about the political, economical and social aspects embedded in the standard. It is not about sheer power or headcount, both because voting is the territory of governments rather than that of civil society, and because building an alternate indirect representation system on a global scale is, for the time being, a foolish idea: even if you could put a meaningful fraction of the worldwide users of the Internet in the same virtual pool of voters, they would not be able to discuss among themselves and express an informed consensus. But this is not a problem, since “astroturf” participation damages everyone, not just users; it is in the interest of the standardization entity itself to ensure that the biggest and purest sample of smart users participates in the works – otherwise, the results will be bad standards that no one will bother to implement.

At the same time, it is the duty of regulatory authorities to ensure that standardization processes are accountable, transparent and democratic, as no standard can be considered open if it is not designed and maintained through such kind of processes. For this to happen, it is necessary to flesh out the principles approved by the WSIS, getting to agreement on a measurable set of criteria; non-binding multi-stakeholder venues such as the Internet Governance Forum may be the best hosts for this activity. Then, a systemic review of the existing standardization venues should be embraced.

¹ “Tunis Agenda”, U. N. document approved at the World Summit on Information Society – Paragraph 29
² “Tunis Agenda”, U. N. document approved at the World Summit on Information Society – Paragraph 52