Anthropic breaks ranks to back state AI safety bill
Anthropic just broke ranks with OpenAI and Google to support a Massachusetts bill requiring catastrophic risk audits every four months. While its rivals warn of fragmented regulation and vague testing standards, Anthropic is leaning into state oversight, a move that could set a major precedent for US AI policy.
Anthropic just broke ranks with OpenAI and Google over state-level AI regulation.
A Massachusetts Senate bill would require frontier AI developers to hire independent evaluators every four months to assess catastrophic model risks. Tucked inside a broader economic development package, the measure won immediate praise from Anthropic's state policy head, Cesar Fernandez.
A four-month audit splits the giants
Google and OpenAI are strongly pushing back. In letters sent August 18–19, OpenAI executives warned that state-by-state mandates will lead to fragmented oversight. Google raised alarms over unclear testing standards that leave developers guessing.
State-level policies have a habit of going viral across the country, much like Illinois's earlier annual audit requirement did.
Why it matters
Anthropic's move looks like a calculated flex. Backing mandatory third-party checks every four months doubles down on its safety-first brand identity while driving up compliance friction for rivals.
We think OpenAI and Google have a fair point, though. Juggling fragmented rules across different states creates operational chaos, and four-month audit cycles could crush smaller players under massive compliance overhead. The regulatory battleground is shifting fast from Capitol Hill to statehouses, and tech giants are no longer speaking with one voice.
Sources (1)
- theinformation.com(opens in a new tab)Originaltheinformation.com