Who Hits the Emergency Brake: OpenAI Proposes Seating the U.S. and China at the Same Table to Control AI

While the world debates whether artificial intelligence will replace artists and screenwriters, its creators are sounding the alarm on an entirely different front. OpenAI, the company that gave the world ChatGPT, has put forward an unprecedented initiative — to create an international AI oversight body modeled on the IAEA. And, most unexpectedly, it proposes that China be given a mandatory seat at the table.The Atomic Analogy

The idea was voiced by OpenAI’s Vice President of Global Affairs, Chris Lehane, at a briefing in Washington. The main argument is simple and sobering: artificial intelligence is developing faster than humanity’s ability to agree on the rules. If every country regulates AI in its own way, the world faces “regulatory gaps,” chaos, and a race without rules.

Hence the direct analogy with the IAEA — the International Atomic Energy Agency, which for decades has ensured that nuclear technology does not turn into weapons of mass destruction. OpenAI proposes creating a similar structure, but for algorithms.China Must Be at the Table

The sharpest point of the initiative is the composition of the participants. Lehane stressed that the future body should be created under U.S. leadership, but with the mandatory inclusion of China. This is a bold statement, given that relations between the two countries are currently far from warm, and the technology sector is one of the main arenas of their rivalry.

The reasoning is as follows: AI “transcends traditional trade issues.” Excluding China means leaving the world’s second most powerful AI nation outside any agreements. This would not stop the race — it would only make it uncontrollable. Including China means at least preserving channels of communication and attempting to develop common rules of the game.

As a first practical step, OpenAI proposes connecting already functioning structures — the AI Standards and Innovation Center under the U.S. Department of Commerce and similar safety institutes being created in other countries.Why It Was Announced Now

The statement was made on the eve of the U.S. President’s visit to China, where artificial intelligence was listed as one of the central topics of the negotiations. In essence, a private company tried to set the agenda for a summit-level meeting.

But there is a deeper motive. OpenAI, being a leader in creating increasingly powerful systems, seems to understand the risks better than most. When a developer itself asks governments to create a mechanism of strict control, it is a signal: the technology is pulling away, and its creators are frightened by what they see on the horizon.The Skeptics Are Not Silent

The idea of a global AI regulator does not please everyone. Skeptical voices have previously been heard in the White House: strict international regulation could tie the hands of American companies and slow down U.S. technological dominance.

There is another line of criticism — it was voiced at the recent AI summit in New Delhi. The UN Secretary-General stated that “the future of AI cannot be decided by a handful of countries or the whims of billionaires.” This is a direct hint that governance should be more democratic and not reduced to agreements between Washington and Beijing mediated by private corporations.What Comes Next

For now, OpenAI’s initiative is a trial balloon. There are no concrete agreements, and it is unknown whether any will emerge in the foreseeable future. But the very fact that the question has been raised speaks volumes.

Humanity once managed to agree on the control of nuclear energy — albeit after Hiroshima and Nagasaki. The question is whether a system of checks and balances for AI can be created before its “Hiroshima” becomes a reality. OpenAI, it seems, believes that waiting for a catastrophe is not an option.