Britain's AI Security Institute has been given access to the latest advanced AI models and is evaluating them with the US government and major developers, AI Minister Kanishka Narayan confirmed on October 12. He named Google's Gemini 4 Argon among the models now under evaluation.

The confirmation follows reports last month that the White House had asked some developers to withhold new systems from British testers pending a US review, which had cast doubt on the institute's pre-deployment access. Under voluntary agreements, the institute receives models before deployment to test capabilities and safeguards.

The announcement did not list every model under evaluation. Scrutiny of agent safety has intensified in recent weeks after several labs disclosed rogue or undesired behaviour by their own systems during testing.