The technology is coming to the NHS whether the law is ready or not. That is the uncomfortable message from the man who runs Britain's healthcare regulator.
Lawrence Tallon, chief executive of the Medicines and Healthcare products Regulatory Agency, told the BBC that AI will soon be used routinely within the NHS — and that the country needs new laws to govern it.
It is a rare thing for a regulator to publicly ask for more powers. When it happens, it usually means the gap between what is already happening and what is legally covered has grown too wide to ignore.
The Regulator Asking for Its Own Rulebook to Be Rewritten
The MHRA is the body that decides whether a medicine, device or diagnostic tool is safe enough to reach patients in the UK. Its approval is the gate between a laboratory and a hospital ward.
Tallon's argument, as put to the BBC, is not that AI is dangerous. It is that the legislation the MHRA works within was designed for a different era of medical products — one where a device did what it was built to do, and kept doing it.
AI systems do not behave that way. They learn, update and shift. A tool approved in January may not be the same tool by December.
Why a Learning System Breaks a Fixed Approval System
Traditional medical regulation assumes a static product. You test it, you approve it, you monitor it. The logic holds because the thing being approved does not change on its own.
AI in healthcare upends that assumption. A model trained on one dataset, then retrained on another, can improve — or quietly degrade. Its behaviour can drift in ways that never show up in the original trial.
That is the regulatory problem in one sentence: the approval was for a version of the tool that may no longer exist.
For patients, this is not an abstract legal debate. It decides whether the software reading your scan, flagging your risk score or suggesting your treatment is being checked against the version actually running in your hospital.
What This Means for Patients Walking Into an NHS Appointment
Most people will never be told that AI was involved in their care. That is precisely why the rules matter.
AI is already embedded in parts of NHS work — imaging analysis, waiting-list prioritisation, administrative triage. The direction of travel Tallon describes points to far deeper involvement: clinical decision support, early diagnosis, and tools that influence what a doctor does next.
The question patients should be able to ask — and currently cannot easily answer — is simple: who signed off on this, and against what standard?
The Gap Between Ambition and Legislation
The UK has spent years positioning itself as a leader in health AI, with the NHS frequently described as a uniquely rich dataset for research. That ambition has run ahead of the statute book.
Post-Brexit, the UK has been reshaping its medical device regime independently of the European Union, which is phasing in its own AI Act. That divergence creates both freedom and risk: Britain can move faster, but it can also fall behind on safeguards.
Tallon's intervention suggests the regulator believes the current framework is not sufficient for what is already on the horizon.
Confirmed Facts vs What Remains Unclear
Confirmed: The MHRA's chief executive has publicly stated that new laws are needed for AI in healthcare, and that the technology will soon be routine in the NHS. This was said in an interview with the BBC.
Unclear: What those laws would specifically contain, who would draft them, what parliamentary timetable exists, and whether the government has committed to legislating. None of this has been confirmed in the available material.
Speculation, clearly labelled: Any claim about when new rules might arrive, or what they would mean in practice for specific AI products, is inference — not fact.
Where the Real Risk Sits — and Where It Doesn't
The case for tighter rules is straightforward: faster approval without adaptive oversight could let flawed tools reach patients at scale, and harm would be hard to trace.
The counter-argument is equally real. Over-regulation could slow the adoption of tools that genuinely improve diagnosis and reduce waiting times — a trade-off that matters in a health service under sustained pressure.
There is also a commercial dimension. Companies building health AI need to know what standard they are being held to. Regulatory uncertainty is its own kind of risk, and it tends to favour large incumbents who can afford to wait.
A Pattern Playing Out Across Every Regulated Industry
The UK is not alone in this. Financial regulators, transport authorities and data watchdogs are all confronting the same structural problem: rules written for fixed products, applied to systems that change themselves.
Healthcare raises the stakes because the consequences are measured in patient outcomes rather than market conduct. What happens with the MHRA will likely become the template for how Britain handles AI in other high-risk sectors.
What Readers Should Take From This
If you are a patient: nothing changes today, but the direction is clear — AI will be part of your care, and the safeguards around it are still being written.
If you work in the NHS: expect evolving guidance rather than a single settled framework, and expect the compliance burden to shift as the rules firm up.
If you are building or investing in health AI: regulatory clarity is now a live variable. Watch the MHRA's public statements closely — they are signalling what they will ask for.
What Happens Next
The immediate next step is political, not technical. A regulator can identify a gap; only government can close it.
Watch for three signals: whether the Department of Health and Social Care responds formally, whether any draft legislation or consultation appears, and whether the MHRA issues interim guidance to cover the period before new laws arrive.
Until then, the NHS will keep adopting AI under rules its own watchdog has now publicly described as inadequate.
Our Take
This story is easy to misread as a warning about AI. It is better understood as a warning about timing.
The technology is not waiting for the law, and the regulator knows it. Tallon's intervention is a signal that the MHRA would rather shape the rules than be overtaken by them — and that the window to do so is narrowing.
For a health service that runs on public trust, that window matters more than any single algorithm.
Frequently Asked Questions
What has the MHRA said about AI in healthcare?
Its chief executive, Lawrence Tallon, told the BBC that AI will soon be routinely used in the NHS and that new laws are needed to regulate it properly.
Is AI already being used in the NHS?
Yes, in limited areas such as imaging analysis and administrative triage. Tallon's comments indicate much broader routine use is expected.
Why are current laws considered inadequate for AI?
Existing medical device rules assume a product stays the same after approval. AI systems can learn and change over time, which makes fixed approval less reliable.
What should patients take away from this?
Nothing changes immediately. But AI is set to become a normal part of NHS care, and the rules governing its safety are still being developed.
[NEWS_SCHEMA] {"@context":"https://schema.org","@type":"NewsArticle","headline":"UK Needs New Laws for AI in Healthcare, Says MHRA Watchdog","description":"MHRA chief executive Lawrence Tallon tells the BBC that AI will soon be routine in the NHS and that existing laws are not ready for it.","image":"","datePublished":"","dateModified":"","author":{"@type":"Person","name":"Newsroom Desk","url":"","sameAs":[]},"publisher":{"@type":"Organization","name":"","logo":{"@type":"ImageObject","url":""}},"mainEntityOfPage":{"@type":"WebPage","@id":""},"articleSection":"Health Technology","keywords":"AI in healthcare UK laws, MHRA AI regulation, NHS artificial intelligence, Lawrence Tallon, medical device regulation UK"} [SOURCES] BBC — BBC News interview with MHRA chief executive Lawrence Tallon --- **Editor's note on sourcing:** This article is based solely on the headline and original story summary provided. No additional high-confidence sources were available at the time of writing. All statements attributed to Lawrence Tallon reflect the original report; no quotes, dates, statistics or official commitments have been invented. Sections marked as unclear or speculative are labelled as such.
