The National Commission into the Regulation of AI in Healthcare has delivered its long-awaited report, setting out a framework for deploying artificial intelligence tools within the National Health Service in a manner described as “safe, fast, and trusted.”
Commissioned by the Medicines and Healthcare Products Regulatory Agency (MHRA), the report concludes that the existing regulatory framework for healthcare AI regulation is fundamentally inadequate, having originally been designed for conventional medical devices rather than the dynamic, evolving nature of AI technologies.
44 Recommendations to Reshape the Framework
Drawing on feedback gathered from more than 12,000 patients, clinicians, and AI developers, the Commission has produced 44 recommendations aimed at updating the regulatory landscape so that AI can become an integral part of healthcare delivery across the United Kingdom.
MHRA chief executive Lawrence Tallon acknowledged the scale of the opportunity, noting that the independent commission has set out “in more detail than ever before” how AI should be governed in healthcare settings.
“We have heard clearly about the opportunities of AI to speed up and improve care, and to release more of clinicians’ time for the essentially human aspects of their work with patients,” said Tallon. “We will only realise those opportunities if we have a modern, dynamic regulatory framework that accelerates safe adoption, protects patients, and commands public and professional confidence.”
Staged Authorisations: The ‘L Plate’ Approach
Among the most notable of the 44 recommendations is the proposal for staged authorisations for healthcare AIs — a concept drawn directly from the learner driver model, where new drivers operate under L plates before earning full independent authorisation. Under this approach, new AI tools would be tested in controlled environments before being rolled out more broadly across NHS settings.
New models would also be introduced with tight controls and close supervision, with full authorisation only granted once real-world safety has been adequately demonstrated. This represents a considered shift away from one-off approvals toward a model of continuous, lifecycle monitoring of AI tools — with rapid withdrawal pathways in place should any system’s efficacy diminish or safety concerns emerge over time.
The report was authored by regulatory science expert Professor Alastair Denniston of the University of Birmingham, alongside Professor Henrietta Hughes, Patient Safety Commissioner for England.
Accountability, Transparency, and Human Oversight
The report places considerable emphasis on accountability for AI developers. It recommends that liability laws and systems be upgraded so that developers can face penalties if their AI tools cause harm, with tougher enforcement mechanisms introduced alongside an updated adverse event reporting system that specifically accounts for AI technologies.
Patients, the report states, must have the right to know when and how AI is actively being used in their care. Information on AI models must be made publicly accessible through a searchable database.
Crucially, the Commission concludes that AI should augment — and never replace — human oversight, with final clinical judgements remaining firmly in the hands of healthcare professionals at all times.
“Any new innovation — including AI — needs to earn its place and earn our trust,” said Professor Denniston. “One message came through clearly: people are open to AI improving their care, but only if it is safe, overseen by humans, and if they know when it is being used. These recommendations are designed to make that trust possible.”
HSSIB Welcomes the Commission’s Findings
The Health Services Safety Investigations Body (HSSIB), an arm’s-length body of the Department of Health and Social Care responsible for investigating patient safety concerns across the NHS in England, welcomed the report, stating that it reflects the need to balance innovation with patient safety, public confidence, and effective oversight.
HSSIB recently launched its own investigation into ambient voice technology (AVT) — a category of AI being rapidly deployed across the NHS for tasks such as summarising clinical conversations and drafting correspondence. That investigation was initiated following the identification of potential safety issues and concerns that existing reporting and monitoring systems may not be adequately equipped to detect errors, hallucinations, or data privacy risks associated with these AI tools.
“As this work progresses, we will carefully consider the Commission’s recommendations and the wider conversations taking place across the healthcare system about the safe use of AI,” said HSSIB interim chief executive Dr Rosie Benneyworth. “By improving our understanding of both the opportunities and the risks associated with these technologies, we can help support their safe adoption in ways that benefit patients, staff, and the healthcare system.”
The Commission’s report marks a considered and evidence-backed effort to move healthcare AI regulation in the United Kingdom toward a framework that is both fit for purpose and capable of inspiring the trust of patients, clinicians, and developers alike.
