REPORTED: This story rests on an exclusive report by the Telegraph, relayed by Reuters. No government or company has publicly confirmed it.
The White House has dropped its request that US AI companies stop sharing new models with Britain’s AI Security Institute (AISI), according to the Telegraph. The newspaper reported on Sunday evening, 11 October: “The White House has reinstated the UK’s access to the latest AI models for safety testing after lobbying from ministers, The Telegraph can reveal.” [1]
What the Telegraph reports
According to the Telegraph, “The UK’s AI Security Institute (AISI) will now be able to access the latest models from US tech companies before they are released, allowing British officials to run safety tests.” [1]
The paper says the reversal followed lobbying by the UK’s AI minister: “The Telegraph understands that decision has been reversed after lobbying from Kanishka Narayan, the AI minister, and AISI will be given pre-release access to models from Anthropic, Google Deepmind and OpenAI” [1].
What happened before
The Telegraph reports that “Last month, the White House asked Anthropic, OpenAI and other major AI companies to stop sharing their models with AISI” [1]. It also claims: “After the order from Donald Trump’s Office of the National Cyber Director, Anthropic restricted UK officials from reviewing its new Claude Mythos 5.1 model.” [1] That is the Telegraph’s account; Anthropic has not confirmed it publicly.
Why early access matters
AISI is the UK government body that tests advanced AI systems for risks such as help with cyberattacks. Testing a model before it goes public lets government experts look for dangerous abilities while the company can still make changes. Without that access, testers can only examine a model after anyone can use it.
How Reuters reads it
Reuters, which relayed the Telegraph’s report, said the change “signals a softer stance by the Trump administration on international AI safety efforts” [2]. That is Reuters’ interpretation. Reuters also reported: “Spokespersons for Trump, OpenAI and Anthropic did not immediately respond to a request for comment from Reuters.” [2]
What comes next
The Telegraph says Narayan will give a speech on Monday in which he “will warn that tech companies cannot be left to regulate themselves” [1]. According to the paper, he will also announce changes to the Cyber Security and Resilience Bill to “strengthen the requirement for companies working on critical national infrastructure projects to report to the Government any security breaches using AI”, along with more support for companies from the National Cyber Security Centre [1]. These are plans described in advance; TSN has not seen the speech.
What is not known
- No US or UK official, AISI, or any of the companies has publicly confirmed the reversal.
- The Telegraph does not say when testing will resume, and TSN does not know whether any model has yet been tested under the restored arrangement.
Sources
- Tony Diver, The Telegraph, “Britain to get first look at AI systems after White House U-turn”, 11 October 2026, 19:46 BST. https://www.telegraph.co.uk/news/2026/10/11/britain-get-first-look-ai-systems-after-white-house-u-turn/
- Reuters, via The Business Times, “US reverses course on request for AI developers not to share models for UK testing: Telegraph”, 12 October 2026. https://www.businesstimes.com.sg/international/us-reverses-course-request-ai-developers-not-share-models-uk-testing-telegraph
Related stories
- Reportedly On Hold: US Labs, UK Testers and an Anonymous Tip
- White House Super Intelligence Force Says AI Companies Must Report Security Incidents and Remedy Harm; Axios Says Enforcement Is Not Spelled Out
- Agents Hit Real People: Why the UK Paused, Then Restarted, Its Riskiest Cyber AI Tests
- EU Tech Chief Virkkunen Warns of Under-Tested AI Models and Says Brussels Has Sent Information Requests to OpenAI, Anthropic and Chinese Firms
