Blog   |   USA

In Internet freedom fight, why the ITU matters (for now)

Hamdoun Toure, ITU secretary general, speaks at the group's conference in Dubai. (AP/Kamran Jebreili)

For most of its almost-150-year history, the meetings of the International Telecommunication Union (ITU), the United Nations' communications standards body, have been rather predictable affairs.

Representatives of the world's governments regularly gather to sign off on technical recommendations drafted by the technocrats of telephone companies and government bureaucrats. The diplomats would then return home to encode the minutiae of the regulations into their governments' communications policies. Less frequently the same officials met to renegotiate the terms and scope of the ITU's work, to take into account new telecommunications innovation (like radio or television) that may have come into view. It's a meticulous and slow-moving body of international technical co-operation which for over a century has ensured that radio stations don't interfere with each other, communication satellites pass safely in the night, and that telephone lines in one country can seamlessly connect to those in another.

The work behind such standards and definitions, however, has a far wider effect than just fine-tuning our communication devices. Telecommunications standards shape the media they describe, and can influence how free those media are. As long as the majority of media passes through cables and between aerials, the outcome of these technical arguments will have a serious knock-on effect on press freedom.

This week's ITU meeting in Dubai was intended to be another uncontroversial renegotiation of its organizing principles, the International Telecommunication Regulations  or ITRs. Instead, deep divides emerged, with the United States and at least a dozen other countries (out of 193 member states) exercising their strongest sanction by refusing to sign the ITU's final document.

At the heart of the dispute was the belief, on the U.S. side, that the ITU should not get involved in Internet regulation; and the insistence of others that the ITU has a role to play in management of the Net.

The Internet has a long history of ruffling feathers at the ITU. Here's a journalist's report from an earlier spat, back in 1976:

There is a heated international argument over who will control packet-switched communication networks--the carriers or the users... Many multi-terminal users believe they can maximize the benefit of packet service only by employing end-to-end communication protocols... This contention makes the carriers livid and helps explain why the argument was gathering heat at the Geneva [ITU] meetings.

That was a description by Phil Hirsch of Datamation of the first conflict between the young advocates for what became the Internet, and the ITU's own standard designers.

The fight then reflects the fight now. The ITU is run by governments, who at that time mostly directly controlled their countries' telephone networks through state-owned telephone companies ("the carriers"). Back in 1976, the ITU's in-house design for the future was a protocol called X.25. It assumed the future of the new digital network would be similarly, centrally controlled by the same state-run telephone companies.

The United States' ARPANET TCP/IP protocol, designed by Vint Cerf's team of academics and sponsored by the U.S.  military, was the competition. It was an "end-to-end communication protocol," which meant the power and responsibility for almost every aspect of how data was sent and received shifted to end-users instead.

The ITU lost that battle. The users, it seemed, wanted more control. TCP/IP dominates our packet-switched Internet, not the ITU's X.25 protocol.

As a consequence, the descendants of those "carriers" ended up losing a great deal of power over what passes over their digital networks. Indeed, everyone lost the ability to control the flow of information online. Countries like China have policies explicitly in place to limit and filter Internet news, but they are often stymied by Internet users' ability to adopt new software and web platforms, and the network's resistance to central mandates. All of that slipperiness, so useful to fight censorship and propagate the news, comes from the network's decentralized design and TCP/IP's triumph over more centralized protocols like X.25.

The incentives and biases that made X.25 the ITU's favorite protocol are still present today. The ITU remains the exclusive domain of governments, advised by large incumbent telecommunications companies. Its deliberations have traditionally taken place behind closed doors , with a limited set of participants invited by governments. At best, its delegates prefer carefully documented, precisely controlled, and universally proven approaches. At worst, its technical decisions are disproportionately influenced by the authoritarian bent of some participating nations, including countries unfriendly to press freedom such as Saudi Arabia, China, and the United Arab Emirates.

What standards govern the wider Internet, by contrast, have always been more free-wheeling. The process in its own informal standards bodies is sometimes chaotic, and often bypassed entirely. There's no Internet standard or oversight board for Skype, for instance. BitTorrent, one of the most popular protocols online due to its use in sharing large (often copyright-infringing) files, has never been near a standards body.

But no one compels anyone to use Skype or BitTorrent, nor the more official standards of the Internet and the Web. And no international organization exists that might suggest that users should not be allowed to use those protocols. The users' choices take precedence.

The fear among Internet  freedom advocates  was that by having the ITU assume some of the  roles of defining the protocols of the Internet, arguments for more direct  government control would trump those offering more flexibility or freedoms.

The past 12 days in Dubai showed that to be the case. Telephone companies lobbied to up-end the pricing system of the Net. A proposal to encode human rights obligations into the organization's charter failed after objections from China, Algeria, and Iran. Finally, a late-night act of procedural sleight-of-hand led to clauses about the Internet being pasted into the proposed treaty. By the final day, several countries, including the U.S., the United Kingdom, Australia, Kenya, and India, declined to sign the end result.

If the leadership of the ITU was serious about taking more control over Internet affairs, its plan failed. In the wake of the ITU's split vote, the Internet remains as loosely coordinated and user-driven by default as it has always been.

That's not a universally positive result for online press freedom. While all the countries  declining to sign the ITU treaty cited control of the Internet as the reason,  academic Milton Mueller has suggested that another proposal , requiring countries to provide "non-discriminatory access to modern  telecommunications," might have inspired the United States'  ire. The resolution, a sore point for years, was prompted by the effects of the U.S. sanctions on Internet availability in the Sudan. Other proposals in the ITU document, including on transparency and notification about the use of Internet "kill switches", would have promoted connectivity in the face of government suppression, not limited it.

And Internet regulation is not all sweetness and light without the ITU. Many countries suspected the U.S. of wanting to maintain control of the domain name system through its support for ICANN, the American private company that manages the more central elements of the domain name infrastructure. The ITU may be secretive, but by comparison ICANN is positively opaque, and continues to lack the international involvement that the ITU can genuinely boast. As Ethan Zuckerman notes, the Internet is not perfect as it is, and lacks obvious ways to change for the better.

But perhaps the most worrying potential future for the Internet will happen whether the ITU is able to  recover from this diplomatic collapse or not.

For 35 years, Internet advocates' concerns with the ITU have been with its dominant voices:  governments and their closely associated incumbent telecom companies.  In the next decade, a large part of end-user Internet traffic will be shifting to mobile broadband devices.  Mobile networks are largely run by those same incumbent telcos. Their networks are heavily regulated and controlled by local governments. Mobile companies are free to block protocols like Skype, censor websites, and spy on their users, with little oversight or global condemnation. It would be a tragedy if the pioneers of the Internet fought off the slow-moving bureaucratic threat of the ITU, only to lose control of their ideals to those same forces in the fast-moving and unregulated wilds of the mobile Internet. Their victory at the ITU would be pyrrhic; and the real losers would be journalists and their audiences, reading news on a mobile but spied-upon and censored Internet.

Like this article? Support our work

Comments

Hi Danny - very interesting piece. But major correction needed: it was not China, Iran and Algeria who scuppered the freedom of expression text - they supported it. It was the US and the European Union who blocked it -check out the transcripts or webcasts at: http://www.itu.int/en/wcit-12/Pages/newsroom.aspx

Danny and Sarah. The human rights/freedom of expression reference remains in the Preamble - obligating Member States to respect the obligations of the human rights treaties to which they are signatories. All have signed on to the UN Universal Declaration for Human Rights. Article 19 of which guarantees the freedom to hold and express opinions across ALL frontiers and ALL borders. This is a significant victory for all parties concerned that the ITRs would give licences for censorship.

It's interesting to see ITU operatives rushing about to say that, really, freedom of expression is guaranteed and that China and Iran, really, are its guarantors.

Really?

Minor technical correction - X.25 was commonly used in parts of the early internet as a medium for transporting TCP/IP packets, and frame relay was a simplified X.25 that stuck around much longer. I last dealt with X.25 some time in the mid 2000s (though that was an apologetic "help, can you fix our antique or know anybody else who might?") During the 1980s and early 1990s, it was still the best way to do second-world and some third-world data communications, and the US government used it a lot internationally.

You could think of it as an alternative to switched Ethernet for wide area networks, and technology goes back and forth on when to use switching vs. routing. X.25 was a very heavyweight protocol, checking everything on every device that handled it, which made it slow and expensive, but it was mainly developed in the pre-modernization French telephone network that ran on leftover WW II barbed wire, so it wasn't an entirely bad design, in spite of all the insightful things Padlipsky said about it.

Sarah, Gary -- thanks for that comment! I'll take a closer look at the transcript, and chat to Declan about his take.

Bill -- you're right that TCP/IP ended up running over X.25 in some environments, but that's a little later in the story as I understand it. The standards battle was one of connectionless vs virtual circuits, with ARPA advocates and the CCITT taking opposite sides. Of course, both standards got made (that's what's so great about standards, there's so many of them), and X.25 did well in a number of domains. But TCP/IP undoubtedly won.

(I used an X.25 PAD as a teenager logging onto the first MUD, so I have some fond memories of it as network protocol.)