Elon Musk proposes peer review for frontier AI models in interview

Elon Musk thinks the biggest AI companies should be checking each other’s homework. In an interview with The Economist published on July 23, the xAI chief executive proposed that leading AI labs submit their most advanced “frontier” models to a peer review process before releasing them to the public. The timing here is not accidental. July 2026 has turned into something of an AI arms race, with xAI’s Grok 4.5, OpenAI’s GPT-5.6, and Anthropic’s Fable 5 all hitting the market in roughly the same window. Musk emphasized the importance of giving competitors sufficient time to evaluate substantial new models in order to safeguard public welfare. Musk referred to Grok 4.5 as “Opus-class,” a nod to the capability tier that places it among the most advanced systems currently in existence. He also advocated for regular safety and security meetings among major AI firms, essentially creating an ongoing dialogue rather than a one-off review. The practical question is whether competing labs will actually open their most valuable intellectual property to rivals, even under controlled conditions. Investors should watch for whether OpenAI or Anthropic respond to this proposal publicly, because the...