UK regulator calls for fresh legislation to oversee AI in the NHS

The Medicines and Healthcare products Regulatory Agency (MHRA) has warned that the United Kingdom’s existing legal framework will struggle to keep pace with the rapid rollout of artificial‑intelligence tools in the National Health Service. Lawrence Tallon, the agency’s chief executive, told the BBC that AI‑driven diagnostics, triage systems and treatment‑planning software are set to become commonplace in hospitals and primary‑care settings, and that new legislation is required to protect patients and maintain public confidence.

According to BBC News, Tallon emphasized that the current suite of regulations – largely designed for conventional drugs and medical devices – does not adequately address the unique risks posed by algorithms that learn and evolve after deployment. He argued that without clear rules on transparency, bias mitigation and post‑market surveillance, the NHS could face unintended harms or legal challenges.

The push comes as the NHS accelerates pilots of AI applications ranging from radiology image analysis to predictive models for chronic disease management. Early trials have shown promise in speeding up diagnoses and freeing clinicians from routine tasks, but they have also highlighted concerns about data privacy, algorithmic bias and the difficulty of auditing “black‑box” systems. Health‑tech firms are eager to scale these solutions, and the government has earmarked billions of pounds for digital transformation, making regulatory clarity a pressing issue.

Britain’s regulatory landscape already includes the Medicines Act, the Medical Devices Regulations and, more recently, the UK’s adaptation of the European Union’s AI Act – a set of standards aimed at high‑risk AI systems. However, critics note that the UK has yet to translate these guidelines into enforceable law specific to healthcare. Tallon’s call mirrors similar demands from professional bodies such as the Royal College of Physicians, which have warned that unchecked AI could erode clinical accountability.

If new statutes are introduced, they are likely to require AI developers to submit detailed safety data before NHS adoption, establish continuous monitoring mechanisms, and grant patients the right to explanations for automated decisions. Such measures would align the UK with emerging global norms and could reassure clinicians wary of relying on opaque technologies.

The MHRA has indicated that it will work with the Department of Health and Social Care to draft the necessary legislation, with an eye toward implementation within the next two to three years. In the meantime, the agency plans to issue interim guidance for NHS trusts deploying AI tools, aiming to balance innovation with patient safety.