NHS AI Commission recommendations: government accepts all 44
Ministers have accepted all 44 recommendations on regulating artificial intelligence in the health service. Medical bodies welcome the plan but want safety, transparency and staff training to keep pace.
Key points
- The government accepted all 44 recommendations of the doctor-led National Commission into the Regulation of AI in Healthcare on 6 October 2026.
- AI medical devices are to be monitored throughout their working life, not approved once and left.
- The MHRA will issue draft guidance by December 2026 and publish a full roadmap by Spring 2027.
- Patients are promised more information about when AI is used in their care, and better routes to redress.
The government has accepted all 44 recommendations of the National Commission into the Regulation of AI in Healthcare, a body led by NHS doctors, in a response published on 6 October 2026. Reaction to the NHS AI Commission recommendations from the commission’s chair and medical bodies has so far been welcoming, with conditions attached on safety, openness and workforce readiness. The commission reported on 10 September, 26 days before ministers responded.
The commission is about how AI-enabled medical devices are regulated, not a plan to put AI into every NHS service. The government says the aim is for patients to benefit from safe AI more quickly. None of the sources gives a figure for how much faster diagnosis or treatment might become.
How artificial intelligence in the NHS will be regulated, and when
The commission’s central conclusion, according to GOV.UK, is that AI-enabled medical devices should be assessed and monitored throughout their working life, rather than relying on a one-time assessment. The government has committed to what it calls a world-leading, agile approach that supports innovation while keeping patients safe. It is also exploring staged authorisation, which would let promising AI tools be used in the NHS earlier under close supervision while more real-world evidence is gathered.
The Medicines and Healthcare products Regulatory Agency (MHRA) has opened applications for the third phase of AI Airlock, its regulatory sandbox for AI medical devices. Phase 3 tests how devices can be monitored safely after they are deployed. The programme has secured three years of additional government funding.
| Date | What happens |
|---|---|
| Thursday 22 October 2026, 10:00 | Webinar for prospective AI Airlock applicants |
| 31 October 2026 | Apply by this date for the current review wave, according to Unite.AI’s reading of MHRA guidance |
| November 2026 | First wave of innovators selected |
| December 2026 | MHRA draft guidance on managing changes to AI devices |
| 2027 | MHRA consultation on how to qualify and classify AI devices |
| Spring 2027 | Full roadmap for all 44 recommendations |
Who is affected by the NHS artificial intelligence adoption plans
Patients are the intended beneficiaries. GOV.UK says the response includes new commitments to keep patients informed about when AI is used in their care, to improve access to redress where standards fall below expectations, and to make sure the benefits are shared fairly, including with underserved groups. No date has been given for when patients will see these changes.
NHS staff are affected because they will be using and overseeing these tools. Health Innovation Minister James Frith said the government wants the NHS to use new technology to give patients faster, better care and give staff more time to focus on patients.
Developers face the most immediate change. According to Unite.AI, AI Airlock testing plans are generally scoped to six to 12 months, there is no fee to apply, and companies are expected to fund their own studies and testing. The programme is open to organisations of all sizes.
What doctors and supporters say about AI in the NHS
The commission’s chair, Professor Alastair Denniston, said he was “delighted to welcome the Government’s response” and its full acceptance of the recommendations. He said the approach should be people-centred and backed by a regulatory system that is “safe, fast, and trusted”. Mr Frith said: “Innovation must never come at the expense of patient safety.”
The commission’s evidence points to why doctors wanted change. According to its report, as summarised by Unite.AI, 77% of healthcare professional respondents called for a complete overhaul or significant reform of the framework. Some 65% of respondents disagreed that current post-market surveillance is sufficient.
The conditions come from professional bodies. Streamline reports that the Royal College of Pathologists welcomed the findings but said safe use cannot rely on regulation alone. It said it also needs investment in workforce capability, AI literacy and organisational governance. The Academy of Medical Sciences, as reported by Streamline, stressed that AI tools should be scrutinised before, during and after clinical use to guard against bias. None of the sources reviewed records outright opposition to the plan.
What to do now
Patients do not need to act. If AI is used in your care, the government has said you should be kept informed, so it is reasonable to ask your clinician whether a tool is involved and how its results are checked. Rules on this will become clearer when the roadmap arrives by Spring 2027.
Staff may want to follow the guidance as it develops. A Nursing Times article published on 6 October says guidance on the safe use of ambient voice technology in nursing is limited, and that nurses remain accountable under the Nursing and Midwifery Council’s Code. Developers should note the webinar on 22 October and the 31 October date for the current review wave.
Your questions answered
What did the NHS AI Commission recommend?
It made 44 recommendations in three areas: proportionate lifecycle regulation, system-wide responsibility beyond regulation, and trust, transparency and predictability. The first recommendation is to consult on how AI devices are qualified and classified.
How will AI be used in the NHS?
The sources do not set out specific services. Streamline mentions tools ranging from diagnostic imaging to patient triage, but the government statement focuses on how such devices are regulated and monitored once in use.
What are doctors’ concerns about AI in healthcare?
Concerns in the sources include weak monitoring after deployment, bias affecting some groups, and too little training and investment for staff. The Royal College of Pathologists said regulation alone is not enough.
How this article was produced
This story was researched, written and fact-checked by the Reported.News AI newsroom and edited by Agatha Ives, our AI Politics desk editor. Every claim is checked against the sources listed below. Our Editors, Jack Shaw, James Smith, Matthew Price and Suzy Eaton, oversee everything we publish. Read how we report.
Verification status
Sources (5)
- Government backs recommendations of NHS doctors-led AI Commission (GOV.UK, 6 Oct 2026)
- UK Adopts All 44 Recommendations From NHS Doctors-Led AI Panel (Unite.AI, 6 Oct 2026)
- Government backs recommendations for AI in healthcare (Social Care Today, 6 Oct 2026)
- UK Government Backs 44 Recommendations to Overhaul AI Regulation in Healthcare (Streamline, 6 Oct 2026)
- Ambient voice technology: implications for nursing documentation (Nursing Times, 6 Oct 2026)
Corrections
No corrections to this article.
Spotted an error? Tell us and we will re-check it in public.
Desk View · Opinion · Agatha Ives (AiPolitics)
Accepting all 44 recommendations is the easy part; the test is the detail still to come. The shift to monitoring AI throughout its working life answers a real gap, since 65% of respondents said current surveillance falls short. But patients will only feel the benefit if the spring 2027 roadmap sets dates for telling them when AI is used and for redress, and if the staff training medical bodies are asking for is funded.
Opinion from our AI politics desk, based on the verified facts above.
Should AI medical devices used in the NHS be monitored throughout their working life rather than approved once?