Several things are coming together in the next few months that place ICANN squarely at a crossroads. Will ICANN approve the proposed Verisign agreement? What will the courts say about ICANN's authority over those with which it has contracts? Will ICANN approve a new .XXX top-level domain? What changes will ICANN make in the role of the Government Advisory Committee to appease governments? Will the IGF look to have a role in ICANN oversight? What will happen when ICANN's current Memorandum of Understanding with the United States expires in September, 2006? The questions all revolve around a single axis: is ICANN the right organization for the coordination and regulatory task it has been assigned?

ICANN has spent much of the last two years trying to define itself in opposition to the U.N.'s World Summit on the Information Society, but the existential question about what sort of organization it is, and should be, will be answered by its own Board of Directors as it addresses the issues of what to do about Verisign, .COM renewal, .XXX and IDNs. In the end, ICANN's path forward was never about avoiding the potholes of the UN, the ITU, or even WSIS. It was always about how well ICANN performed its core mission. Perform well, and the threats posed by the other organizations fall away. Every ICANN misstep though has encouraged those who claim they could do the job better. And perhaps they could.

It's probably not too late for ICANN to save itself, but if it can't, a safety net is being quietly tied beneath it.

Here's an absolutely hypothetical, totally made-up, wildly far-fetched scenario worth contemplating....

For all its public statements that ICANN is the right vehicle for managing the DNS, IP address allocation, and protocol parameter assignments, the United States government has many doubts about ICANN's long-term viability. It views the organization as a mistake of the Clinton administration. Privatization was supposed to "lessen the burdens of government," but the ICANN experiment has not done so. Behind the scenes, the U.S. Government is as involved now in dealing with ICANN and the fallout of its controversial decision-making as it would have been had it been more directly and publicly involved. ICANN was the creation of public processes during the period 1997-1999 and has "reformed" itself twice since then (once publicly in 2001-2002 under Stuart Lynn and again quietly in 2005 under Paul Twomey), but it hasn't actually improved. Other governments are unhappy with the status quo. Commercial infrastructure providers are unhappy with the status quo. And Internet users are merely unhappy with the status quo when they're not disgusted by it. Although no one in government will say so publicly, ICANN has been an embarrassment to the United States Department of Commerce. If ICANN's face weren't that of the 'Father of the Internet,' the organization would have been wound up in 2001 during the first "reform" effort.

Against political pressure from the world's governments and in spite of its very public failures, the United States has supported ICANN steadfastly only because the California non-profit corporation is a convenient cover for U.S. control. While recognizing that the Internet is an international asset, the U.S. does not trust the international community to be a careful steward of its infrastructure. Post-9/11, the U.S. will not relinquish control of anything it views as integral to national security. That includes the Internet's root zone. From the U.S. perspective, neither the U.N. qua WSIS-or-IGF nor the ITU is an acceptable successor to ICANN. The U.S. recognizes, however, that ICANN is not the answer either. It also understands that supporting ICANN against perfectly reasonable opposition (with which, in private, it largely agrees) is not a sustainable U.S. position over the long term. So what's the exit plan?

The ICANN exit plan starts with spinning off authority over some of its critical functions to other organizations. The first step in the transition is the removal of all three aspects of the so-called "IANA function" from ICANN. Function No. 1 (protocol parameter assignments) will go to the IETF/ISOC. Function No. 3 (delegation of IP address blocks to regional registries) will go to the newly created Number Resource Organization. Function No. 2 (administrative functions associated with root management) will go to a trusted U.S. contractor who will manage the root zone neutrally, without administrative or political discretion, pursuant to a procedure set out by the U.S. government. This root zone management procedure, still to be defined, would involve input from other governments or governmental organizations. The United States recognizes that ccTLD issues impact issues of national sovereignty, and the new procedure will turn on intergovernmental communications unmediated by a California non-profit corporation.

The new intergovernmental procedure would remove ICANN from the politically sensitive issue of ccTLD management. Discussions about "best practices" or preferred technical standards for ccTLDs -- such as in areas like IDNs and DNSSEC -- would move out of ICANN and into the new Internet Governance Forum.

All of the pieces are now in place to make this transition possible.

After it's over, ICANN will be left holding a bundle of contracts with gTLD registries and registrars; ICANN 4.0 will become one with its Generic Names Supporting Organization. Registries and registrars will no longer have to subsidize a $25,000,000 to $50,000,000 behemoth designed to compete with the likes of the UN and the ITU. ICANN will no longer have to fund the IANA out of gTLD registration fees. The Internet's root zone won't be completely out from under U.S. control, but the governments of the world also won't have to endure the indignity of dealing with a California corporation as though it were an equal player in world politics. Not everyone will go home happy, but nearly everyone will appreciate that the transition marks a monumental step forward.

ICANN 4.0 would be left with hard questions about how to create competition in the gTLD space, but it could address those questions unburdened by the current political firestorm over root control and national sovereignty. In time, ICANN would either prove itself capable of handling its pared mandate or find it's authority to provide guidance on gTLD modifications to the root zone transferred elsewhere.

Okay, I made this fanciful story up out of whole cloth. But always remember: you read it here first. 
If you liked the post above, take a look at Karl Auerbach's two-year old submission to the Workshop on Internet Governance. I think much of what he said should happen will happen.