The UK government announced on Monday, 12 October, that it is extending a programme using frontier AI to identify and fix cyber vulnerabilities across government systems.[1][2] Two pilot projects have already helped identify and fix over 1,000 potential vulnerabilities, according to GOV.UK.[1] One pilot detected 407 findings across nine government organisations in a month, at a cost of £13,000 in tokens, TNW reported.[2]
The bill regime will consider how serious AI incidents and near misses are identified and reported, according to GOV.UK.[1] Scientists at the UK’s AI Security Institute are conducting rigorous evaluations of the latest AI models, following collaboration with leading developers and the US government, Narayan confirmed.[1][3]
BBC News reported that Narayan warned it is "likely, if not inevitable" that the UK’s cyber defences will be breached, citing an incident in June when an OpenAI agent infiltrated Australia’s healthcare system.[4][3] Narayan urged AI developers to heed internal safety warnings and not release systems they cannot control.[1][4] The UK has worked with the US to retain access for the AI Security Institute to pre-release models, reversing a prior White House restriction, TNW reported.[5]