OpenAI and Anthropic back binding UK AI safety rules; Google DeepMind and Meta decline

At a UK parliamentary hearing, OpenAI and Anthropic said voluntary AI safety pledges are no longer enough and that new laws should be introduced.

Anthropic's head of Europe, Liam Booth-Smith, told the committee that "the rules should be binding" and that "basing everything on voluntary commitments for frontier risk is not a good long-term plan."

Google DeepMind and Meta declined to back new legislation, according to Wired's report. UK officials said the AI Security Institute's pre-release testing remains voluntary and that it cannot stop a company from releasing a model it judges too dangerous.