The UK government announced new preparations for AI cyber risks on October 12. Its official statement says public-sector pilots have helped identify and address more than 1,000 potential vulnerabilities, and that the programme will be extended.
BBC reporting on October 12, carried by AOL, corroborated the announcement. Minister Kanishka Narayan said Britain had worked with the US to retain pre-release access to powerful models for its AI Security Institute. That statement does not specify every model or a date when any earlier restriction ended.
The government will examine stronger protections through the Cyber Security and Resilience Bill, including serious AI incidents and near misses. Those proposals are not newly enacted requirements. Work on responding to autonomous-agent incidents is underway; it does not guarantee that cyber attacks will be prevented.