White House lifts its ban on UK safety testing of new AI models
Washington had asked Anthropic, OpenAI and other companies to withhold new models from the UK's AI Security Institute. It changed course after lobbying by UK AI minister Kanishka Narayan.
The White House has reversed its ban on the United Kingdom getting access to the newest AI models for safety testing, LBC reported on October 12, citing The Telegraph. Washington had earlier asked major AI companies, including Anthropic and OpenAI, to keep their models away from the UK's AI Security Institute. Because of that request, Anthropic stopped UK officials from reviewing its new Claude Mythos 5.1 model. According to The Telegraph's reporting, the US changed course after lobbying by Kanishka Narayan, the UK's Minister of State for Artificial Intelligence. LBC said Narayan is also set to warn about growing risks from AI. The decision matters because the UK institute is one of the main government bodies that tests frontier models before or around their release. The ban cut it off from at least one new model. Restoring access brings back a channel of independent government evaluation for models built in the US. The supplied report does not say when the original request was made, what conditions apply to the restored access, or when UK testing of Mythos 5.1 will begin.