OpenAI CFO Backs Slower AI Development as Safety Debate Intensifies in 2026

2 articles · Updated · CNBC · Sep 15

Sarah Friar said OpenAI would slow frontier AI work if researchers judge it necessary, adding that she would make business decisions around any safety-driven pacing.
OpenAI’s CFO said she is not worried about AI agents using weapons of mass destruction to kill, but stressed the company is taking safety and alignment “really seriously.”
A former Anthropic researcher’s warning that AI could “kill us all by the end of the decade” has pushed safety concerns to a fever pitch in Washington and across the industry.
Dario Amodei, Elon Musk and Sam Altman have all backed calls for slower development, with Amodei arguing Saturday that labs should move more cautiously to reduce the risk of something going seriously wrong.
The renewed debate leaves IPO timing uncertain for frontier labs: Altman has called a 2026 OpenAI listing “ill-advised,” while Friar told employees in August the company expects to go public in 2027.