Anthropic CEO calls to ‘slow the pace’ of AI development amid safety concerns

Anthropic chief executive Dario Amodei wrote in a blog post Saturday that the development of artificial intelligence needs to be slowed so teams can implement necessary risk-prevention safeguards. He framed the request as a way to ensure safety measures keep pace with advancing model capabilities. Amodei cautioned that even if the rate of capability improvements is reduced, advances will still feel fast, and he argued for prudent management of that rapid progress. The blog post urged deliberate steps to make sure safeguards are put in place rather than allowing capability gains to outstrip protections. The message reflects a call from a leader inside a prominent AI company for industry-wide attention to safety timing and risk mitigation. By recommending a slower development tempo, Amodei emphasized creating the conditions to design and deploy preventative measures before capabilities become harder to control. Anthropic’s statement adds to an ongoing conversation about how to balance innovation with precaution as AI systems grow more powerful. The company framed the slowdown as a practical step to enable effective risk management rather than a halt to research and development.