Question

To ask the Secretary of State for Digital, Culture, Media and Sport, what arrangements are in place for a serious incident involving autonomous AI systems, with reference to the Written Statement of 7 September 2026 on Artificial intelligence update, HCWS314; which Minister and department would lead the response; when the proposed government agentic AI incident response capability will become operational and what interim arrangements apply; whether cross-government exercises have tested scenarios involving coordinated AI agents compromising multiple organisations; and what assessment the Government has made of the adequacy of its powers to secure the restriction or suspension of a model’s use where a serious risk is identified.

The Government is taking significant action to strengthen the UK’s resilience to agentic AI incidents. This includes the establishment of a specialist capability to investigate and respond to incidents. This new capability, initially announced in the Written Statement of 7 September 2026 and allocated £40 million in the Budget on 28 October 2026, will be called the National Agentic Security Unit and is being established by the National Cyber Security Centre (NCSC) and the Department for Digital, Culture, Media and Sport (DCMS). The Government will provide further details in due course. In the meantime, we have robust and well-established procedures in place to respond to a wide range of incidents, including cyberattacks.

The Cyber Security and Resilience Bill will introduce new national security direction powers. Where strict legal tests are met, the Government can direct regulated entities to take specific action where a security or operational compromise creates a national security risk. In an AI loss-of-control scenario, this could include steps to isolate systems, restrict data flows or suspend automated actions. Where further powers are needed to keep the British public safe, we will act.