A Forbes commentary argues a broad US AI slowdown could weaken US leadership and urges stronger safeguards. On Sept. 23, Dario Amodei and Sam Altman sought testing standards and incident reporting at the UN Security Council; the White House opposed a pause.
On Sept. 12, Anthropic CEO Dario Amodei proposed slowing the rate at which frontier models improve, not stopping training, and warned rogue agent swarms could cause internet-scale damage within months. Sam Altman and Elon Musk endorsed the proposal that day.
Peter Thiel said the risk of losing control was real and that even a 5–10% chance was serious; he argued a genuine pause would require enforceable cooperation among Washington, Beijing and every lab. The commentary distinguishes delaying a model a lab cannot safely deploy from halting the sector, advocating targeted safeguards and investment in defensive AI, cybersecurity, biosecurity and infrastructure.
