UK Parliament Summons AI Giants Over Safety Regulations

The chair of the Business, Innovation, Science and Trade Committee, Liam Byrne, has summoned senior representatives from OpenAI, Anthropic, Google DeepMind, and Meta to a parliamentary hearing on October 13. The inquiry examines whether the United Kingdom's voluntary AI safety regime is fit for the future, specifically questioning if pre-release safety testing should be legally mandatory and if regulators should have the power to block dangerous models.

This summons follows concerns that voluntary cooperation is insufficient, highlighted by Anthropic releasing a new model without prior testing by the AI Security Institute. The committee is also investigating personal accountability for developers and legal liability for foreseeable harm caused by frontier models.

In a related move toward transparency, OpenAI announced it will allow third-party organizations to conduct technical safety assessments during the training and evaluation phases of its models, rather than only before launch. This follows a promise by CEO Sam Altman to provide independent evaluators with employee-level access and the right to publish findings. OpenAI is currently in discussions with groups such as METR and Redwood Research, while competitors like Anthropic have already embedded evaluators from Accenture. OpenAI also faces adversarial testing requirements under the European Union's Article 55 for models with systemic risk.