Elon Musk Calls for Joint AI Safety Reviews Among AI Labs: Willing to Put Aside Differences with Altman and Supports China's Participation in AI Collaboration

In a recent interview with The Economist, Elon Musk argued that the world’s leading AI labs should create a shared safety review process before releasing advanced new models. His view is that as frontier systems improve at an accelerating pace, companies should not operate in isolation. Instead, they should give one another time to examine major models for safety concerns ahead of launch and report serious risks to government authorities in both Washington and Beijing.

Musk’s proposal goes beyond coordination among U.S. firms. He said any meaningful safety framework should also include Chinese AI developers, reflecting his belief that AI oversight will be more effective if it is built on broader international cooperation rather than limited to one country or one bloc.

Cross-lab safety checks before major model releases

At the center of Musk’s comments is the idea that AI labs should not rush powerful systems into public use without outside review from their peers. He suggested that before a new frontier model is released, other leading labs should be given a window of time to assess possible safety issues. If material risks are identified, those concerns should then be communicated to the relevant government bodies.

The underlying message is that the rapid pace of capability gains in advanced AI demands more structured coordination. In Musk’s view, informal competition alone is no longer enough; there should be a mechanism for labs to exchange warnings and flag dangers before highly capable systems are deployed.

• AI labs should allow peer review of frontier models before launch.

• Potential safety problems should be reported to authorities in the U.S. and China.

• The system should include Chinese companies rather than be limited to American participants.

A rare opening toward Sam Altman

Musk also addressed whether he would be willing to work with OpenAI CEO Sam Altman despite their long-running tensions. His answer suggested that, if necessary, he would be prepared to do so. He said that if circumstances ultimately required a conversation, then that conversation would happen, adding that he would be willing to set aside personal disputes for the benefit of the world.

That statement stands out because Musk and Altman have been at odds for years over the direction of OpenAI and the broader trajectory of the AI industry. Musk has repeatedly accused Altman and OpenAI of moving away from the organization’s original nonprofit vision in favor of commercial interests. Their conflict has spilled into lawsuits, and Musk has frequently criticized Altman publicly on social media.

Against that backdrop, his latest remarks were widely seen as an unusual sign of flexibility—at least on the issue of AI safety, where he appears to believe cooperation may matter more than personal rivalry.

Opposition to efforts aimed at blocking China’s AI progress

Musk also made clear that he does not support attempts by U.S. companies to ban the use of Chinese AI models. In his assessment, such restrictions are unlikely to meaningfully slow China’s progress in artificial intelligence. He argued that even if the United States continues trying to limit China’s access to the most advanced technologies, Chinese AI developers will still keep advancing.

On the issue of AI chip supply, Musk said he believes China will eventually make its own breakthroughs, including through the development of domestic lithography equipment. That view is especially notable given the broader technology tensions between the United States and China.

A global approach to AI governance

Taken together, Musk’s comments point to a consistent theme: he sees AI safety as a problem that cannot be handled effectively through national silos alone. Rather than trying to wall off China from the conversation, he appears to favor a governance model that recognizes the global nature of AI development and the need for communication across borders.

His proposal for joint safety reviews, his willingness to engage even with longtime rivals if necessary, and his rejection of strategies designed to simply contain China all reflect the same broader position. As AI systems grow more capable, Musk is arguing for a framework built on coordination, risk reporting, and international participation.