OpenAI Backs US-Led Global AI Governance Body With China

OpenAI has thrown its weight behind the creation of a U.S.-led global artificial intelligence governance body that would include China as a member, a proposal floated hours before President Donald Trump arrived in Beijing for high-stakes talks with Chinese President Xi Jinping. The initiative signals a significant push for international cooperation on AI regulation at a time when geopolitical tensions and technological competition between the two superpowers are at a peak.

A Bold Proposal at a Critical Moment

Chris Lehane, OpenAI’s Vice President of Global Affairs, told reporters on May 13 that the company would support establishing an international AI governance framework modeled on the International Atomic Energy Agency (IAEA). The IAEA, founded in 1957, sets global safety standards for nuclear energy development and includes both the United States and China as member states.

According to Bloomberg, Lehane suggested the framework could be built by linking the U.S. Commerce Department’s Center for AI Standards and Innovation with AI safety institutes being developed around the world. “AI, in some level, transcends a lot of the prevailing or traditional trade type of issues,” Lehane said. “There is an opportunity to really start to build something up globally, and have countries around the world, including China, potentially participate.”

Fox Business reported that Lehane made the comments at a briefing at OpenAI’s Washington offices, just as Trump’s delegation — which included Nvidia CEO Jensen Huang and more than a dozen other U.S. business leaders — was arriving in Beijing for the first state visit by a U.S. leader to China in nine years.

The Mythos Catalyst

The OpenAI proposal comes against the backdrop of growing concern over frontier AI capabilities, particularly following Anthropic’s announcement of its Claude Mythos model in early April. According to BBC News, Mythos Preview has already found thousands of high-severity vulnerabilities, including some in every major operating system and web browser — one vulnerability had been present in a system for 27 years.

Anthropic’s Project Glasswing, which gives 12 major tech companies including AWS, Apple, Microsoft, Google, and Nvidia access to Mythos for defensive security testing, has demonstrated that frontier AI models now possess cyber capabilities that can surpass all but the most skilled human experts. The UK’s AI Safety Institute confirmed Mythos is a very powerful model, though it noted its biggest threat would be against poorly defended systems.

The G20 has since requested answers from Anthropic on the risks posed by Mythos, with the company agreeing to brief global financial regulators and central banks, as reported by India Today. Bank of England Governor Andrew Bailey said regulators are “having to look very carefully” at what Mythos could mean for cybercrime risk.

Administration Skepticism and Political Headwinds

Despite OpenAI’s push for international governance, it remains unclear whether the Trump administration would welcome global guidelines that include China. White House officials have previously indicated they would reject the idea of worldwide governance over AI technology. The administration has been preparing an executive order on AI cybersecurity that emphasizes voluntary rather than compulsory pre-deployment model review.

Former White House AI and Crypto Czar David Sacks, who previously shaped AI policy in the Trump administration, offered a nuanced perspective. Speaking on Fox Business’s “The Claman Countdown,” Sacks said, “I do think that there are things that may be in our common interest, and it’s worthwhile to explore having those conversations.” However, he also cautioned that “the fact is, we have to still protect from against each other. So I think it’s going to be a little bit limited in terms of what we can achieve there.”

As reported by The New York Ledger, Sacks also downplayed the need for strong federal regulation of AI, arguing that “the real issue is not what the American labs do. It’s the fact that Chinese models and other models that other actors could train are gonna have advanced cyber capabilities within the next six months or so.”

The Commercial Dimension

One of the underlying drivers of OpenAI’s proposal is the commercial tension between U.S. AI companies and their Chinese counterparts. U.S. firms, including OpenAI and Anthropic, have complained that Chinese developers are using the outputs of cutting-edge American models to produce rival systems at a fraction of the cost and with fewer guardrails. A governance framework could potentially address intellectual property concerns while maintaining market access.

Before the U.S. delegation departed for Beijing, officials said they would explore the possibility of opening a new channel of communication with China to regularly discuss AI issues — a significant shift that may have been accelerated by the emergence of Mythos and its demonstrated capabilities.

What’s Next

The key question is whether the Trump administration will embrace or reject OpenAI’s proposal. The outcome will have significant implications for the future of AI regulation, US-China relations, and global technology governance. While the IAEA model offers a precedent for international cooperation on dual-use technologies, the political reality of US-China rivalry — spanning trade, technology transfer, and human rights — creates formidable obstacles.

OpenAI has also called for the U.S. to require government researchers to evaluate cutting-edge models in classified settings, suggesting the company sees a role for both domestic oversight and international coordination. As Sacks put it, “I think the point here is for the two sides to start talking, to establish an initial dialogue and just to see how the Chinese are thinking about this.”

Whether that dialogue leads to a functioning governance body — or remains limited to cautious conversations — will depend on whether both Washington and Beijing can find common ground on an issue that, as Lehane noted, “transcends a lot of the prevailing or traditional trade type of issues.”