UK Commission Sets Out a Framework for AI Healthcare Regulation

The UK now has a roadmap for AI healthcare regulation. On 10 September 2026, the National Commission into the Regulation of AI in Healthcare published its recommendations for how the country should oversee artificial intelligence in medicine, balancing faster access to new tools against patient safety and public trust.

The commission was set up by the Medicines and Healthcare products Regulatory Agency (MHRA) in September 2025. It brings together expertise across health, regulation and AI, with contributors from the United States and Singapore. Over a year it heard from more than 12,000 people: patients, the public, clinicians, healthcare leaders, industry and technology developers. The MHRA describes this as the largest engagement of its kind ever run in the UK on the regulation of healthcare technology.

The message that came back was consistent. People are open to AI playing a bigger role in their care, but on clear conditions: strong safety standards, meaningful human oversight, and transparency about when and how AI is used. AI is already at work in the NHS, spotting strokes and skin cancers early and freeing up clinician time through voice-enabled tools. The commission wants to extend those benefits to far more patients, and to make the UK the best place to build, test and use AI in healthcare.

Four recommendations for AI healthcare regulation

The report turns on four changes to how AI-enabled medical devices reach patients and are supervised once they do.

Staged authorisations

The first is staged authorisations, much like the L-plates a learner driver carries. A new model would be deployed under close supervision with tight guardrails, and would earn fuller approval only after it has shown real-world safety and performance. The aim is to give UK patients early access to promising models while keeping risk under control at each step.

Continuous monitoring, not a one-off approval

The commission also wants to move past the single, point-in-time approval. Because AI can behave differently in different settings and can change after deployment, it recommends continuous real-world monitoring across a device’s whole working life. The logic mirrors real-world evidence in medicines reimbursement: a device should be judged on how it performs in routine use, not the day it was cleared.

Public access to safety information

The report calls for the public to be able to search for information about the safety of specific AI-enabled medical devices, including any adverse incidents. It points to the MHRA’s interactive Drug Analysis Profiles as a model. Patients have also told the commission they want to know when AI is being used in their care, and the recommendations set out a proportionate way to keep them informed.

Stronger enforcement powers

Finally, the commission recommends stronger powers for the MHRA, including enhanced enforcement, so the regulator can act decisively when an AI system falls short of the standards patients expect.

The public’s conditions

A deliberative study by the Health Foundation, the commission’s research partner, worked with 78 members of the public in Cardiff, Milton Keynes and York between March and April 2026. It sets out five principles for regulating AI in health care. Accuracy came out as the public’s top priority, alongside human oversight as a condition of use, regulation proportionate to the risks involved, and a guarantee that AI does not lead to worse care for any group.

The commission’s work lands as regulators elsewhere tighten their own rules. The EU’s AI Act has set strict requirements for high-risk systems in health, and AI healthcare policy has expanded quickly across the United States and other markets.

What happens next

The government and the MHRA will now consider the recommendations, with a formal response to follow. Professor Alastair Denniston, the NHS consultant who chairs the commission, framed the task as one of earning trust: “These technologies need to show that they are safe, effective and bring benefits to patients, staff and the wider NHS, without leaving people behind.”

Professor Henrietta Hughes, the commission’s deputy chair and Patient Safety Commissioner for England, said patients want to know when AI is used, that it supports rather than replaces clinicians, and that there is clear accountability when things go wrong.

Professor Neil Lawrence, chair of the technology working group and DeepMind Professor of Machine Learning at the University of Cambridge, noted that generative AI can behave differently in different circumstances and evolve after deployment. “We can’t rely on a single point of approval and assume the job is done,” he said.

Jennifer Dixon, chief executive of the Health Foundation, said the real test will be whether the NHS has the capacity, skills and systems in place to implement and monitor AI applications safely and effectively at the scale now needed.

Source: Medicines and Healthcare products Regulatory Agency