OpenAI CEO Sam Altman on Sunday backed calls to “slow” AI development to prevent
frontier models from outpacing human control. He echoed Anthropic CEO Dario
Amodei, saying a shift toward safety will carry costs but is warranted to
maintain confidence in U.S. firms’ responsible development of powerful AI.
Altman wrote that “any U.S. competitive pressure should not be an excuse for
recklessness, nor allow capabilities to outrun alignment and monitoring; when we
talk about ‘slowing’ we do not mean ‘stopping’.” The remarks follow rare public
agreement from Amodei and Elon Musk. Altman told staff earlier this month OpenAI
is willing to slow frontier progress to better understand risks and confirmed
the company will not IPO in 2026 while prioritizing safety.
