Anthropic and OpenAI split on halting risky frontier AI in California
Original title: Anthropic and OpenAI diverge on who should pull the plug on risky AI
Anthropic wants courts able to suspend models; OpenAI wants SB 53 to require developers to use shutdown controls.
In submissions after Governor Gavin Newsom’s September 18 executive order N-9-26, Anthropic and OpenAI both back independent safety assessments and broader incident reporting, but differ on stopping risky models. Anthropic’s October 15 filing says the attorney general should be able to seek a court order suspending deployment, internal use or further development when near-term catastrophic harm is found. OpenAI’s October 20 letter asks to amend Senate Bill 53 so frontier developers must maintain and use shutdown controls when no lesser step addresses the risk. The order itself imposes no new requirements; recommendations are due November 16.
- Anthropic proposes risk reports at least every six months and a state-pooled fund for evaluators.
- OpenAI wants annual compliance audits plus technical assessments, coordinated with NIST.
- OpenAI says open-weight copies on external systems cannot reliably be switched off by the developer.
- Both prefer targeted controls, such as stopping a training run or isolating an agent, over one kill switch.
- Countries
- United States
From the original article
3 min read at csoonline.com
OpenAI and Anthropic agree that frontier AI systems need stronger oversight, but they are proposing sharply different approaches to stopping models that pose serious risks: Anthropic wants courts empowered to suspend dangerous systems, while OpenAI argues developers should be required to maintain and use effective shutdown controls.
csoonline.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.