Anthropic, OpenAI, Google Quietly Discussed an AI Safety Standards Body

Anthropic, OpenAI and Google have quietly discussed creating an artificial intelligence safety standards body, according to a report by The Information. The discussions point to interest among major AI developers in a possible shared framework for addressing safety issues as increasingly capable systems are deployed across consumer, business and research applications.

The report does not establish whether the companies reached an agreement, what authority such a body might have, or which standards it would seek to set. Any industry-led organization would likely face questions about membership, governance, transparency and how its work would interact with government rules and existing technical standards efforts.

AI safety has become a central policy and industry issue as companies release powerful models that can generate text, images, software code and other outputs. Standards can provide common approaches for evaluating systems, documenting risks and developing safeguards, though voluntary industry initiatives are distinct from binding regulation. Anthropic, OpenAI and Google are among the best-known developers of advanced AI models.