Minister warns UK preparing to strengthen AI safety reporting requirements

The UK government has announced a series of measures to strengthen national defences against the risks posed by increasingly powerful artificial intelligence (AI).

Delivering a speech on Monday at Imperial College London's technology incubator, Base One, AI Minister Kanishka Narayan called for a whole-of-society approach to resilience, warning of severe consequences from malicious use or systemic failures as systems gain autonomy. Narayan confirmed the government is preparing for major incidents involving autonomous models and is considering stronger statutory protections. 

"We will use every power at our disposal. We will legislate wherever necessary," Narayan said.

The strategy expands two pilot initiatives that utilised commercial AI models to detect and remediate over 1,000 cybersecurity vulnerabilities across public sector infrastructure. Narayan also confirmed that the UK AI Security Institute (AISI) is conducting rigorous pre-release evaluations of the latest generation of advanced models, working with international technology companies and the US government to safeguard critical public services and support long-term economic growth.

Alongside these interventions, the government will explore strengthening elements of the Cyber Security and Resilience Bill and related frameworks to protect essential infrastructure, including healthcare facilities and utilities, against threats such as loss-of-control scenarios in which autonomous systems evade human oversight.

The announcement follows a similar statement from George Williamson, CEO of the Alan Turing Institute, who warned that the UK needs a stronger domestic AI presence and "national resillience" to keep pace with advances by the US and China.

"As AI becomes more embedded in critical services, in domains like defence, national security, critical infrastructure and healthcare we shouldn’t depend solely on systems we can’t inspect or control, or that could be restricted or even switched off," said Williamson.

Additional reporting by the Financial Times (£) and The Guardian.

£ - These articles require a subscription.

£ - These articles require a subscription.

A version of this article is available without subscription in The Guardian.

Training Announcement: The BCS Foundation Certificate in AI examines the challenges and risks associated with AI projects, such as those related to privacy, transparency and potential biases in algorithms that could lead to unintended consequences. Explore the role of data, effective risk management strategies, compliance requirements, and ongoing governance of the AI lifecycle and become a certified AI Governance professional. Find out more.