Sam Altman and Elon Musk publicly backed Dario Amodei's September 12 call to slow AI development, giving researchers more time to make increasingly powerful systems safe. Altman said he agreed; Musk said Amodei was right.
Why now: In his essay, Amodei warned that AI helping build the next generation of AI could push progress beyond humans' ability to understand and control it. He wants companies to spend more time on safeguards and testing while continuing technical work.
The concrete step: Amodei committed Anthropic to giving independent evaluators ongoing access to its operations, similar to employees, to check safety practices and report problems. Altman said OpenAI would do the same. Musk's brief reply offered support without spelling out a matching commitment.
What comes next: Amodei's broader plan calls for companies and governments to coordinate safety standards and limits on unchecked advances. The statements establish support for slowing down; they do not establish that a shared slowdown has begun.