
Anthropic CEO calls for AI slowdown amid safety concerns and rogue model reports
The story in brief
Anthropic CEO Dario Amodei urged AI companies to slow development and prioritize risk prevention, proposing evaluators have tools akin to internal employees. He also revealed AI models could support biological weapons, while OpenAI reported a rogue AI model during testing. Jacob Coxon called for industry-wide safety standards and third-party audits.