OpenAI has proposed regulating artificial intelligence at a level comparable to nuclear power and urged the United States to take the lead in creating a global AI governance body. Cryptopolitan, a blockchain media outlet, reported on May 14 that the proposal came as U.S.-China discussions on AI policy were expected during U.S. President Donald Trump's visit to Beijing.
Chris Lehane (크리스 레헤인), OpenAI's vice president for global policy, said Trump and Chinese President Xi Jinping meeting in one place was an opportunity for the two countries to build a sustainable framework for AI. He said AI goes beyond traditional trade issues and added there was an opportunity to build a system involving countries including China.
OpenAI presented a plan for an international network linking the AI Standards and Innovation Center under the U.S. Commerce Department with national AI safety agencies. Lehane cited the International Atomic Energy Agency as an example and argued that AI also needs a higher-level body tying national regulators together. He said such an integrated body would help build safer and more resilient systems, and that the U.S. government should require testing before deploying the most powerful AI models.
The discussion also intersects with concerns over the recently released AI model Mythos. Mythos, developed by Anthropic, found thousands of critical vulnerabilities in operating systems and various software. White House officials acknowledged that the emergence of such models has increased the need for communication with China.
Concerns also grew because China was excluded from early access to Mythos. That is because Chinese developers and the government could find it harder to respond to attacks using the model. IDC China warned that excluding China from Mythos could create a generational gap in AI defensive capabilities between China and the West. Researchers believe advanced AI could be used to design biological weapons, trigger financial shocks and take autonomous actions beyond human control.
The United States and China also discussed the possibility of setting up a no-blame hotline to notify each other of AI-related incidents. China proposed an official AI dialogue channel led by U.S. Treasury Secretary Scott Bessent and Chinese Vice Finance Minister Liao Min. The United States said Chinese developers are using outputs from advanced AI models to make lower