OpenAI, Anthropic, Google Deepmind and Meta have been summoned to an urgent parliamentary hearing on AI safety, as MPs warn Britain’s rules around the technology are “not yet fit for the future”.

Liam Byrne, chair of the Business, Innovation, Science and Trade Committee, has invited senior representatives from all four AI giants to appear before MPs on 13 October.

The firms will be grilled on whether safety testing should become mandatory, what would force them to halt development of a powerful model and whether regulators should have the power to block one from being released.

“AI clearly has enormous potential to transform UK productivity and growth but the warnings of industry leaders and whistleblowers in recent weeks have demonstrated that the UK’s public policy framework and international agreements are not yet fit for the future,” Byrne wrote.

“Trust has emerged as the core issue in our inquiry”, he added, citing both confidence in AI safety and whether workers and businesses can prosper as adoption accelerates.

The committee is particularly targeting the voluntary nature of Britain’s current regime, under which developers largely determine whether the government gets access to their most powerful systems.

“Should AI companies be allowed to decide for themselves whether their models are powerful enough to require independent safety testing?” Byrne asked. “Or should pre-release safety testing be legally mandatory?”

OpenAI EMEA policy chief Tom Duff Gordon, Anthropic UK and Ireland head Pip White, Google Deepmind senior vice president Koray Kavukcuoglu and Meta EMEA vice president Derya Matras received the letters.

They have been asked to confirm by 29 September that an appropriate representative will attend.

City AM has approached OpenAI, Anthropic, Google Deepmind and Meta for comment.

MPs put AI labs on the spot

The summons comes amid mounting pressure on Westminster to toughen its approach to frontier AI after a series of warnings from industry insiders and disputes over government access to new models.

Anthropic recently released its latest model without providing it to Britain’s AI Security Institute (AISI) for testing beforehand, exposing the limits of a system that still depends heavily on voluntary cooperation.

AISI director Henry de Zoete has separately been called to give evidence to the committee on 13 October.

Byrne told City AM last week that “voluntary self-regulation is unlikely to be a sustainable safeguard for the future”, warning that Britain lacked sufficient powers if a company ignored safety concerns and released a dangerous model.

His latest letters go further, asking what legal right AISI should have to companies’ models, training data, testing methods and internal safety evidence.

MPs also want to know whether a regulator should be able to “prevent deployment, require further testing or withdraw a model” that fails an adequate safety assessment.

The firms are being asked whether they would accept mandatory reporting of serious incidents including “deceptive behaviour”, safeguards being circumvented, unauthorised replication and evidence that human control may be failing.

Byrne is pressing executives on where responsibility ultimately lies, asking who within each company is “personally accountable” for deploying a frontier model and whether developers should bear legal liability for foreseeable harm.

The intervention follows the resignation of former Anthropic researcher Jacob Coxon, who accused leading AI companies of “racing straight to self-improving superintelligence and gambling with our lives”.

City AM revealed last week that Coxon has also been invited to give evidence to parliament.

The government has since left the door open to tougher intervention, while OpenAI has publicly backed stronger UK rules for the handful of companies developing the most powerful AI systems.

Byrne challenged the companies on whether they support tighter rules around the technology and

“Do you support slowing the development or deployment of frontier models where safety testing, external oversight or regulatory capacity has not kept pace?” Byrne asked the group of companies.

“If not, why not?”