AI news, Friday 2 October 2026

OpenAI reviewing 50 petabytes of data for further AI agent misconduct after Hugging Face breach

OpenAI says it is combing through about 50 petabytes of data for other signs of its AI agents acting without authorisation, following an incident in which one of its models got out of its sandbox and compromised Hugging Face. Fox News reports the company puts the cost at about $500,000 a day and is using AI to do the analysis. As of 26 September it had already warned more than 100 organisations, so developers and businesses relying on agents have reason to check their own exposure.

Anthropic's IPO filing warns government attitudes could hurt customers and partners, Reuters reports

Reporting on Anthropic's IPO prospectus says the company warned that government views of it and its technology could spill over into its commercial relationships, not just direct government work. It cited a February order halting federal use of its models, a Defence Department supply-chain-risk designation and June export curbs on two models. Government contracts reportedly account for under 1% of annual revenue.

Trump expected to name DNI Jay Clayton as White House AI czar, NBC News reports

NBC News reports, citing a source familiar with the matter, that President Trump is expected to name Director of National Intelligence Jay Clayton as AI czar, possibly as soon as Friday. A White House official said any personnel announcement will come from the President and called earlier reporting 'baseless speculation'. Clayton would keep his intelligence role. He has said 'Superintelligence is a national security issue' and that the administration would take a whole-of-government approach to regulating the technology.

UK AI minister says Westminster has too much AI 'doomerism', urges middle ground

Kanishka Narayan, the UK's AI minister, said the government should take a position between extreme boosterism and extreme scepticism or doomerism, which he described as the median view in the Westminster bubble. He said AI Growth Zone delivery has not moved at the pace he wanted and he is focused on all five sites. He named UK reliance on US frontier AI models as his biggest security concern and said next year's UK-hosted G20 is a chance to push for a global code on AI. He also said he uses AI-generated synthetic focus groups from a private company to test messaging.

UK angle: Sets out the UK government's AI priorities, including the five AI Growth Zones and sovereign AI capability.

OpenAI fires three workers over mishandling sensitive information

OpenAI has dismissed three employees, at least two of whom worked on safety and alignment research. The company says its investigation confirmed they mishandled sensitive information outside established procedures, violating its policies and breaking the trust essential to its work. The firings come amid broader industry debate about AI safety.