Elon Musk urges AI firms to hold safety reviews before new releases

Elon Musk has proposed that the world’s leading AI companies should establish regular joint reviews before releasing new technologies, in order to reduce risks associated with rapid AI development.

In his interview with The Economist on July 29, 2026, Musk suggested that executives from firms such as xAI, Anthropic, OpenAI, and others meet weekly or biweekly to discuss safety concerns.

He argued that before any company launches a more advanced AI system, competitors should be given one to two weeks to evaluate it for potential dangers and provide feedback. Musk stressed that governments should not be the first line of review, since they often lack the technical expertise of developers. Instead, peer companies are better positioned to identify risks.

If a company refuses to address serious safety issues flagged by others, Musk said governments could then intervene to halt or regulate the release. He also emphasized that this cooperative framework should include Chinese AI firms, given their rapid progress in the field.

Musk concluded that because AI is advancing at extraordinary speed, such safety discussions should begin immediately rather than waiting months. His proposal reflects growing global concern about how to balance innovation with responsible oversight in the AI sector.