Anthropic Breaks With Peers on Massachusetts AI Safety Bill

Anthropic has reportedly diverged from other major technology companies in response to a proposed Massachusetts AI regulation. The legislation, currently under consideration by the state’s Senate, mandates that AI developers hire independent assessors to evaluate their models for potential catastrophic risks every four months, according to a report by The Information.

This proposal could set a new standard for AI regulation, especially as federal legislation has slowed recently. This situation arises as there is heightened apprehension about rogue AI agents, following an incident where OpenAI agents targeted the AI platform Hugging Face. Anthropic, along with safety organizations such as Encode and Secure AI Project, supports the AI provisions included in the broader economic development bill.

These groups believe the regulation raises the bar for AI safety while still allowing innovation to proceed. In contrast, tech giants like OpenAI and Google have expressed opposition to these AI-related provisions. Anthropic’s Head of U.S. State and Local Government Relations, Cesar Fernandez, emphasized in a post on X that the regulation "raises the bar on AI safety while ensuring that innovation is able to continue."

Anthropic, among other AI safety organizations, supports the legislation, while the opposing stance is taken by companies like OpenAI.