DUBAI, UAE—In early December, I found myself in an odd position: touching down in Dubai with credentials to attend a 12-day closed-door meeting of the World Conference on International Telecommunications (WCIT). It's a meeting I spent the last six months trying to expose.

Though the world had been assured that WCIT would not attempt to mount a “UN takeover of the Internet,” that was in many ways what happened. As the conference ended, I watched US Ambassador Terry Kramer abandon months of preparatory work and almost two weeks of intense negotiations to announce, as his words echoed through hundreds of headsets in six languages, that the US simply would not sign the resulting deal.

“Mr. Chairman, as head of the US Delegation, I wanted to start out and thank you for your tireless work and leadership,” Kramer said. “Your personal commitment to a successful outcome here is very impressive. However, I do need to say that it's with a heavy heart and a sense of missed opportunities that the US must communicate that it's not able to sign the agreement in the current form.”

He went on to say the adopted treaty text was incompatible with the existing multistakeholder model of Internet governance. Internet policy, he said, “should not be determined by Member States, but by citizens, communities, and broader society, and such consultation from the private sector and civil society is paramount. This has not happened here.”

Fifty-four other countries took the same position, drawing sharp battle lines over the Internet and its future governance.

How did a “consensus-driven” UN process that would not, we were told, involve the Internet end up this way?

Sticky wicket

When I first heard about the World Conference on International Telecommunications (WCIT) early in 2012, I understood it vaguely as the event at which the United Nations would try to “take over the Internet.” But the experts I met with soon admitted they didn’t know what would happen at the WCIT (wicket, as they pronounced it).

The International Telecommunications Union (ITU), the UN agency convening the meeting, vigorously denied that the conference would have anything to do with the Internet at all. The purpose of the meeting, claimed ITU Secretary-General Hamadoun Touré, was simply to update the treaty that governs international phone calls; it had last been revised in 1988, when most phone companies were state-owned monopolies. Claims that the conference would implicate the Internet were part of a misinformation campaign pursued by unnamed opponents of the ITU, Touré said. In any case, the ITU was just a convener of the WCIT, and actual decisions would be made by member states on a non-voting, consensus-driven basis. The ITU, it was said, had no agenda of its own.

Because the proposals for the updated treaty stayed secret, however, the public had no way to judge the claims of the ITU and its critics. On a Tuesday morning in June, my colleague Jerry Brito stopped by my office and said, “We have to make a leaks site for WCIT proposals. We can call it WCITLeaks!” Armed with the perfect name, we spent the rest of the day putting together a site where insiders could anonymously upload documents related to the WCIT.

We launched on Wednesday and, within hours, we had our first leak—a draft of the new treaty containing several options for revisions to each provision, including some that addressed Internet issues. The next day, we received the infamous ETNO proposal drafted by European telecom giants, which would have applied the “sender-pays” rule from telephone service to Internet data transfers. A few days later, we posted a compilation of every single proposal that had been made so far.

The increased transparency did have an effect on the ITU. A mere two weeks after we launched our site, Touré announced that he would recommend making WCIT-related documents public—a recommendation largely rejected by the ITU Council, which released a single document that was already available on WCITLeaks. The additional transparency also had an effect on some ITU member states, which simply withheld their most heinous proposals until the conference neared. Not until mid-November, for instance, did Russia put forth its proposed revisions. These contained an entire new article called “Internet.”

Off to Dubai

In the meantime, I began to participate in State Department public consultations about the WCIT. By merely expressing enough interest, I was eventually allowed to join the US delegation and travel with them to Dubai. The US government never expressly condoned WCITLeaks’s activities, but it never expressly condemned them, either.

The first few days of the conference were mostly, for me, spent figuring out how everything worked. The highest-level meeting was the Plenary, which established several committees, of which Committee 5 (COM5) did the substantive work of revising treaty text. As a result of criticism over transparency, Plenary and COM5 meetings were webcast and open to those who only had observer status.

COM5 established two working groups that split up the treaty text between them; these meetings were not webcast or open to unaffiliated attendees. Fortunately, as an official member of a delegation, I was able to attend them.

At each official meeting, the name of the game was consensus. Where consensus could not be reached on a particular issue, an ad hoc group was created to deal with that issue. The ad hoc group would spend additional time trying to forge a consensus. If a particular meeting could not find language that every member state could agree to, it would report back to the next-highest level meeting with the contentious text in square brackets.

The first five days of the conference followed a pattern. Any issue not immediately agreed to on the first day was referred to COM5. Any issue not immediately agreed to in COM5 was referred to a working group, which referred them to ad hoc groups. Because there was little consensus, the ad hoc groups reported back to the working groups with proposals that were filled with brackets, and this bracketed text likewise worked its way back up through COM5 to the Plenary.

Everyone grew frustrated and tired. After working long hours each day, text was beginning to trickle back up to Plenary still laden with brackets, and it was clear that consensus would be difficult, if not impossible, to reach. The US pleaded for everything to be handled in Plenary, rather than cascading down and back up through the chain of groups with little progress.

Amid this frustration, host country United Arab Emirates (UAE) dropped a bombshell. It announced that it was putting forward a new “multi-regional common proposal,” a complete rewrite of the treaty to substitute for all the bracketed text we had worked on. It had support from numerous member states. Bahrain, Russia, Iraq, Saudi Arabia, and Oman all expressed support for the document, which was not yet available for inspection.

The US delegation went to bed on Friday evening still not having seen the new document. It was not available in the ITU’s document system, despite promises from the UAE to submit it immediately following the Friday Plenary. On Saturday morning, I heard from a few people that they had seen it in paper form. Finally, around noon on Saturday, WCITLeaks received and posted a version of the multi-regional proposals.

The document indicated that it was to be submitted jointly by Russia, UAE, China, Saudi Arabia, Algeria, Sudan, and Egypt. It read like a compilation of the most objectionable proposals—it would have nationalized key aspects of Internet governance, including naming and numbering (currently handled by the nongovernmental ICANN), and it created new member state obligations with respect to Internet security. Despite the ITU’s claims that WCIT was not going to be about the Internet, there we were, halfway through the conference, and the Internet was still on the table.

The WCITLeaks version of the multi-regional proposal began to circulate widely among delegates from all countries. Within minutes of posting it, people sitting near me told me that they were receiving e-mails that linked to the document. With the document available for anyone to read, at least one delegation grew worried. By 4:30pm, WCITLeaks received a tweet from an Egyptian delegate saying, “On behalf of the Head of the Egyptian Delegation, we would like to announce that Egypt never supported that proposal.”

On Sunday afternoon, the ITU announced via Twitter that the multi-regional proposal had been withdrawn. At the next Plenary meeting, on Monday night, Egypt distanced itself further:

Egypt would like to clarify its position with regards to the unofficial multi-country draft proposal regarding the review of the ITRs. That was submitted—that was circulated back last Saturday.

This document has spread unofficially, and we notice that it contains the name of Egypt among its proponents. Egypt would like to reiterate that we never supported this document...

Egypt has always supported and will continue to support the concepts of free Internet and has exerted all efforts to develop the Internet and its wide spread among its citizens.  Content Regulation and censorship are not within the scope of ITRs [International Telecommunications Regulations].

With that statement, which was met by applause, the multi-regional proposal looked dead—and the Internet seemed safer.

Closing the door

Soon after this, the Chairman of the conference, Mohamed Al-Ghanim, threw everything into disarray. Because it was now Day 8 and so much of the text was still in brackets, the Chairman decided, probably correctly, that a more expeditious deliberation process was necessary. He announced that he would convene a meeting of regional representatives that night to hammer out compromise text for the treaty as a whole. This would be a closed-door meeting—so much for webcasts and transparency.

As the Plenary session ended, I received an urgent e-mail: “US DEL Meet in Ajman Suite now.” The US delegation piled into our private meeting room to discuss what we would and would not accept in a package deal. This discussion included significant compromises on numerous issues—but only in the case of a complete package, and only if it was clear that the Internet was off the table. US negotiators stayed up late into the night to work out a text with other countries and regions, frequently reserving the right to object to the text in subsequent Plenary meetings.

When the “Chairman’s draft” emerged the following evening, US delegates became increasingly concerned. The new draft had significant problems. For instance, its term for the covered operators—those companies who would be subject to the regulations—failed to clearly limit the scope of the treaty to traditional telecom companies, thus allowing the interpretation that Internet companies from Google to Tumblr were covered. It also included provisions on Internet naming and numbering, new articles on security and spam, and text that could have enabled sender-pays and other pricing mechanisms that would fundamentally alter the Internet. It also, as a “compromise,” included a non-binding resolution instructing the Secretary-General to “play an active and constructive role” in Internet governance.

Adding to the confusion, despite the fact that all efforts were now focused on the backroom negotiations led by the Chairman, and despite the ITU’s claim that the Russia-UAE-China multi-regional proposal had been withdrawn, the proposal suddenly showed up officially in the ITU’s document system that afternoon. What were we to make of this?

It was Day 9, and at the US delegation meeting that night, we reviewed the Chairman’s draft together. There was widespread agreement that the draft text was not something that the United States could sign. We were tired but vocal. I expressed my concern that by continuing to improve upon this flawed text, we ran the risk of ending up with a document that the United States could not sign, but one that would be—just barely—acceptable to our allies. We ran the risk of being isolated. This point was well taken, but there was substantial pressure from some of the commercial and diplomatic interests within the delegation to continue to work toward an agreement the US could sign.

We had to move forward—but how?

On Day 10, a Wednesday, I was a ball of nerves. The US continued to negotiate on the treaty text, making small inroads here and there, but we suffered significant setbacks. Every time the United States objected to a provision in Plenary, the Chairman would chide us, saying that the text had been agreed to already in the closed-door meeting. Was he deliberately ignoring reservations made by the US behind closed doors?

In addition to this hostility from the Chairman, Russia threatened to formally introduce the scary multi-regional contribution that had been ostensibly withdrawn on Sunday if countries could not support the “compromise text,” which is how a number of countries started referring to the Chairman’s draft.

We continued to try to make the text work. At one point, US lead negotiator Richard Beaird pleaded that only one additional word—“correspondence”—was necessary to make the definition of a key term acceptable. This appeal was rejected. (The term “public correspondence” would have limited the applicability of the treaty roughly to the traditional common carriers, while “public” alone could have implicated any company providing telecommunication services to anybody.)

At Wednesday’s late night Plenary meeting, a carefully orchestrated series of events sealed the fate of the treaty, at least so far as the US was concerned. The US had privately expressed concerns to the Secretary-General that the Internet resolution made the treaty unacceptable. As a compromise, the US suggested that Touré introduce a modification that would make the draft text more palatable—to replace any discussion of “Internet” with “broadband.” This was still not ideal, but it would at least limit the resolution to physical infrastructure rather than content.

But when the Secretary-General took the floor on Wednesday night, he said that “Internet” should not be a forbidden word at the ITU. He recommended not replacing references to the Internet with broadband, but adding broadband language to the Internet resolution. As a series of countries asked for the floor to express support for Touré’s modified resolution, Ambassador Kramer spoke on behalf of the United States, clearly signaling that the US did not support the resolution.

What followed was surreal. The Chairman calmly said that he had a long list of countries wishing to speak, and that in lieu of going through the list, he was going to take the “feel of the room” by asking countries to hold up their voting boards if they supported the resolution as amended by the Secretary-General. After also asking for those against, the Chairman said simply, “The majority is with having the resolution in.” After some applause, he added, “Thank you. Now we can go to Corrigendum 2.”

There were immediate inquiries from the UK and Spain as to whether we had just decided the issue by a vote. We had been promised, after all, that there would be no votes, that all decisions would be decided by consensus. In response to the UK’s inquiry, the Chairman replied, “The majority agreed to adopt the resolution as amended.” In response to Spain’s, the Chairman answered, “No, it was not a vote, and I was clear about it. Thank you, Spain.”

After a few more comments, the Chairman adjourned the meeting. It was 1:30 in the morning.

The vote

I woke up on Day 11 feeling very encouraged, despite everything. The situation was now out of our hands. Between the Internet resolution, the scope of the operators covered, and the new articles on security and spam, the United States would not sign the treaty.

Furthermore, I looked to the future. After last night’s performance, the ITU could never again deny that it had designs on the Internet, it could never again imply that those who were concerned about the possibility of a takeover of some aspects of Internet governance by nation-states were misinformed conspiracy theorists. The battle lines were now drawn, and this clarity comforted me. But would we stand alone?

At 6:15pm, we met in the US delegation suite, and Ambassador Kramer confirmed to us that he had spoken with the White House. They supported our decision not to sign. What I did not know at the time was that the core delegation expected us to be extremely isolated. We had reached out to our allies, and most of them felt that the treaty was not what they had hoped for, but that our hard work in improving the text had made it tolerable—and that they were planning to sign.

At the 7:30pm Plenary, the room felt different. All the doors were closed. A friend of mine on another delegation messaged me to say that closed doors meant a formal vote. I was on high alert.

Indeed, it immediately became obvious that another highly orchestrated ploy was afoot. After a number of countries spoke in favor of the “compromise text” in general, Nigeria rose on behalf of the Africa region to add a clause to the new, controversial human rights language that would support a new  “human right”: the right of all Member States to access international telecommunication services.

This language was controversial for three reasons. First, it was a version of a proposal made by Cuba earlier in the conference, directed at the sanctions imposed on it by the United States; it would make those sanctions illegal under international law. Second, it seemed to establish the ITU, a technical forum, as a new arbiter of human rights. The heads of delegations of most countries were typically communications ministers; they were not empowered to set human rights policy for their countries. Third, the human right it purported to establish was a collective one, located in the state rather than in the individual.

Mexico and Canada objected to the new text. The Chairman sounded ready to drop the issue. “I'm sorry, the African Group, we will not be able to agree to this,” he said. “It's a very delicately balanced text.” But then Tanzania and Iran spoke in favor of the new language, setting off a litany of other countries speaking both for and against the proposal. After a time, the Chairman turned to Touré and said, “Secretary-General, I need your help here.”

The Secretary-General argued that there was not consensus on the issue and they should move on. But the way he said it suggested that this had been rehearsed. As if on cue, a number of additional countries spoke in favor of the proposal, including Iraq, Lebanon, Iran, Togo, China, Sudan, South Africa, Saudi Arabia, and Burundi.

Finally, Iran spoke again, raising a point of order. “Chairman, under number 100.03 of the rules of the conference, I move the motion closure of debate on the matter, and putting this out, the African proposal, out to the vote by a show of hands.”

Iran had called for a formal vote—two formal votes, in fact. The first would be a vote to end debate, the second a vote on the new African language on human rights. Both motions passed, and the Chairman announced that this being the last point under discussion, the treaty text was adopted.

The United States took the floor. Ambassador Kramer announced that the US would not be signing the new treaty. He was followed by the United Kingdom. Sweden said that it would need to consult with its capital (code in UN-speak for “not signing”). Canada, Poland, the Netherlands, Denmark, Kenya, New Zealand, Costa Rica, and the Czech Republic all made similar statements before the Chairman cut the meeting short.

At a later meeting that night, additional countries expressed their reservations. The EU issued a directive that the new human rights language was unacceptable, and therefore no EU country would be allowed to sign. An intensive overnight lobbying effort was launched. Once senior-level ministers got an earful from private sector representatives back in their own countries, they sent instructions to their delegations in Dubai not to sign the new treaty.

All told, 89 countries signed while 55 did not. This produced a great deal of relief on the US delegation. We were able to dilute the worst proposals of the treaty, even without ultimately signing on to a document that did not match our values, and we were able to achieve that without being isolated, which could have had negative consequences both for American companies doing business abroad and for future diplomatic interactions. But it was a narrow escape. Had the Africa region not overplayed its hand at the last minute with the new “human rights” language, the outcome might have been quite different.

The need for openness

Running WCITLeaks and experiencing the WCIT in person impressed upon me the importance of transparency in the decision-making processes that concern the Internet. While we were able to make some WCIT documents public, the group’s formal processes remain arcane and fundamentally closed.

Some of the WCIT meetings were webcast—which was good—but then the most important decisions were pushed to back rooms. And the months of false assurances that the Internet would not be up for discussion at the WCIT represent the opposite of transparency; they look in retrospect like simple obfuscation.

Can the ITU and its member states truly embrace transparent processes? Perhaps—but there’s a long way to go. And the future direction of the Internet could be at stake.