The Government has announced plans to “step up national defences” in the face of growing risks from artificial intelligence, including exploring possible changes to the Cyber Security and Resilience Bill.
The announcement came as the AI Minister, Kanishka Narayan, warned that increasingly sophisticated frontier AI systems capable of acting autonomously present significant risks, going so far as to say that it is “likely, if not inevitable” that the UK’s cyber defences will be breached at some point.
To address these risks, the Government has announced a range of measures, including “putting frontier AI to work” to help identify and address vulnerabilities, and exploring with the National Cyber Security Centre and AI Security Institute (ASI) how to respond to incidents involving AI agents. The ASI will also continue to evaluate the latest generation of models, while the Government considers how near miss incidents can be identified, reported, and learned from as quickly as possible.
As for the Cyber Security and Resilience Bill, which begins its Report Stage on 26 October 2026, the Government has confirmed that it is exploring how it and “the wider regime” might be strengthened to “ensure the UK’s essential services such as hospitals and utilities are better protected from emerging AI related risk. This includes AI loss of control risk, where autonomous AI systems evade human oversight”.
The announcement is a cautionary reminder of the cyber security implications of both agentic AI and artificial intelligence more broadly. Emphasising that businesses themselves have an important part to play, the Government directs organisations to its open letter sent earlier this year (discussed here), urging them to strengthen their defences against AI-related cyber threats.
To read more, click here.