NHS doctors set out 44 rules for how AI should be used in your care
What happened
On 10 September 2026, the National Commission into the Regulation of AI in Healthcare published its report. It sets out 44 recommendations for how the UK should oversee artificial intelligence used in health and care. The Commission was set up in September 2025 and is chaired by Professor Alastair Denniston, an NHS consultant who leads the UK's Centre of Excellence in Regulatory Science in AI and Digital Health. Its deputy chair is Professor Henrietta Hughes, an NHS GP and the Patient Safety Commissioner for England. Over a year, the Commission engaged more than 12,000 people: patients, members of the public, clinicians, healthcare leaders, industry and technology developers. It describes this as the largest engagement of its kind ever carried out in the UK on the regulation of healthcare technology. Separate deliberative research run by the Health Foundation involved 78 members of the public in Cardiff, Milton Keynes and York during March and April 2026. Among the recommendations are staged authorisations, so a new AI tool is first used under close supervision and tight guardrails while it proves itself in real conditions; continuous monitoring of AI medical devices throughout their working life rather than a single approval at the start; a way for the public to look up safety information and reported incidents for a specific AI device; a clear expectation that patients are told when AI is used in their care; and stronger enforcement powers for the MHRA so it can act quickly when an AI system underperforms.
Why this matters
AI is already in the NHS. It helps read scans to spot strokes and skin cancers earlier, and voice tools are being used to write up consultations so clinicians spend less time typing. The problem is that the rules we have were mostly written for devices that do not change. A hip replacement behaves the same way in year five as it did on day one. An AI model does not necessarily. It can be updated, it can be applied to a population it was not built for, and its performance can quietly drift as the world around it changes. That is why the Commission's emphasis on monitoring across the whole life of a product, rather than a one off approval, is the substantive part of this report. It is a different way of thinking about safety. There is also a trust argument, and it is not sentimental. If people do not trust how AI is being used in their care, they will avoid services, withhold information, or opt out of data sharing. That makes care worse for everyone, including the people who were never worried in the first place. Getting the rules right is the thing that makes the benefits reachable.
What the evidence actually says
Two things are worth separating: what the Commission found out, and what it recommends. What it found out came from that very large engagement exercise. The consistent message from the public was conditional acceptance. Professor Denniston summarised it by saying people are open to AI improving their care, but only if it is safe, overseen by humans, and they know when it is being used. That is not opposition to AI. It is a set of conditions. What it recommends is, at this stage, advice. A commission report is not law. The 44 recommendations now have to be considered by government and regulators, and some would need changes to how the MHRA works and possibly to legislation. Separately, MHRA regulatory reform amendments were tabled in the Government's Health Bill on 1 September 2026, and the MHRA and Manchester NHS organisations announced a health innovation sandbox on 2 September, so there is real movement around this. But movement is not the same as delivery. It is also fair to say that the evidence on how well AI performs once it leaves the research setting is still thin in places. Tools that do well in published studies do not always do as well in an ordinary busy clinic with different equipment and a different patient mix. That gap is precisely what staged authorisation and lifecycle monitoring are designed to catch.
Practical advice
You do not need to become an expert in medical device regulation. But you do have some straightforward rights and options. You can ask. If you are told a scan was reviewed with AI help, or your consultation was written up by a voice tool, it is entirely reasonable to ask how it was used and whether a human checked the result. Under the Commission's recommendations, being told should become the norm rather than something you have to prise out. Remember that a human clinician remains responsible for your care. AI tools in the NHS are used to support decisions, not to replace the person making them. If something in your care does not make sense to you, say so. Be cautious about health AI outside the NHS. Symptom checker apps and chatbots that are not regulated medical devices are not held to these standards, and some are marketed in ways that overstate what they can do. They are not a substitute for clinical advice. And if you think something went wrong with a device or a piece of health technology, you can report it to the MHRA through its Yellow Card scheme, which covers medical devices as well as medicines. Reports from patients genuinely feed into regulatory action.
What to know
A national commission chaired by NHS doctors has published 44 recommendations on regulating AI in healthcare, after engaging more than 12,000 people over a year. The headline proposals are staged authorisation for new AI tools, continuous monitoring across a product's life, public access to device safety information, telling patients when AI is used in their care, and stronger enforcement powers for the MHRA. AI is already used in the NHS to help spot strokes and skin cancers and to reduce paperwork. The report is advice, not law, and what happens next depends on government and regulators. You can always ask how AI was used in your care, and a human clinician remains responsible for it. Sources: GOV.UK, 'Independent Commission led by NHS doctors sets out blueprint to accelerate safe AI adoption in healthcare', 10 September 2026, https://www.gov.uk/government/news/independent-commission-led-by-nhs-doctors-sets-out-blueprint-to-accelerate-safe-ai-adoption-in-healthcare ; Burges Salmon, 'New UK Commission to set regulatory rulebook for AI in healthcare', https://www.burges-salmon.com/articles/102l7ch/new-uk-commission-to-set-regulatory-rulebook-for-ai-in-healthcare/ ; MHRA, National AI Commission meeting minutes, https://assets.publishing.service.gov.uk/media/694bd3bbecb1863cca9d30f1/National_AI_Commission_2nd_meeting_public_minutes.pdf This article is for general information and does not replace advice from a doctor, pharmacist or other qualified healthcare professional.
Related articles
A pain-relief patch the size of a plaster, controlled from a phone
A research team in South Korea has built a small wearable device that delivers electrical pain relief through the skin, and can be controlled remotely from a sm…
Something dark is hiding in Venus's clouds, and scientists just narrowed down what it can be
Venus has a mystery that has sat unsolved for roughly a century. Seen in ordinary light, the planet is a smooth, featureless cream-yellow ball. Look at it in ul…
Scientists looked a quarter of a proton deep inside an atom - and found something odd
Physicists working on the ALICE experiment at CERN's Large Hadron Collider have managed to look deeper inside atomic nuclei than anyone has before. What they fo…