44 recommendations, accepted in full, and a sandbox that now watches what happens after approval
What happened
On 6 October 2026 the government confirmed it will accept all 44 recommendations made by the National Commission into the Regulation of AI in Healthcare. The Commission, led by NHS doctors, published its blueprint on 10 September 2026. Accepting every recommendation is unusual. Governments more often accept most, reject a few, and quietly park the rest. Alongside the response, the Medicines and Healthcare products Regulatory Agency (MHRA) opened applications for the third phase of AI Airlock, its regulatory sandbox for AI-enabled medical devices. A sandbox is a controlled space where new products can be tested with the regulator watching, before the rules are fixed. AI Airlock has secured three years of additional government funding. Phase 3 focuses on something specific: post-market surveillance, meaning how an AI device is monitored once it is actually in use. A webinar for people thinking of applying is planned for 22 October.
Why this matters
Medical device regulation was built around things that do not change. A hip implant approved in 2020 is the same hip implant in 2026. AI is not like that. Models are retrained. Software is updated. Performance can drift as the patients, scanners or hospitals change around it. A one-off approval tells you how a tool behaved on the day it was tested, which is not the same as how it behaves in your hospital two years later. The Commission's central conclusion was exactly this: AI-enabled medical devices should be assessed and monitored throughout their working life, rather than relying too heavily on a single assessment at the start. That sounds technical. It is really about trust. If an AI tool helps read your scan, you want to know someone is still checking whether it works, not just that it worked once in a trial. The government response commits to a more proportionate, lifecycle-based framework, with the stated aim of letting patients benefit from safe and effective AI sooner while keeping safeguards in place as the technology changes.
What the evidence actually says
It is worth separating what has been decided from what has only been promised. Decided and happening now: AI Airlock Phase 3 is open for applications from developers of eligible AI-enabled medical devices. Its theme is post-market surveillance and lifecycle regulation - testing, with real products, how AI devices can be monitored and managed safely after deployment. Committed with a date: the MHRA says it will issue draft guidance by December 2026 on a new approach to managing changes to AI-enabled medical devices as they adapt and improve over time. Committed for next year: the MHRA will consult on how to qualify and classify AI-enabled devices. That was the Commission's first recommendation. Being explored: new staged authorisation pathways, which would allow promising AI tools to be used in the NHS earlier under close supervision while more real-world evidence is gathered. Still to come: a full implementation roadmap setting out timelines and responsibilities across all 44 recommendations, due by spring 2027, with regular progress updates after that. The government has also committed to strengthening patient involvement in decision-making, improving access to redress where care falls below standard, and making sure the benefits of AI reach underserved groups rather than only well-resourced places. So: one concrete programme open today, one piece of guidance due within months, and a longer list that will be judged on the roadmap due in 2027.
Practical advice
For most people, nothing changes in the GP surgery this week. But there are reasonable things to know. If an AI tool is involved in your care, you are entitled to ask about it. A fair question is whether a human clinician reviews the output and makes the final decision. In NHS practice, AI tools used in diagnosis are generally decision support, not decision makers. Do not assume an app you download yourself carries the same oversight as a device used in the NHS. Regulated medical devices go through MHRA processes. Wellbeing apps often do not, even when they sound clinical. If something about your care goes wrong, the government response specifically mentions improving access to redress. The existing routes - raising it with the provider, then PALS, then the ombudsman - still apply. If you work in a health technology business, Phase 3 applications are open now and the applicant webinar is on 22 October. The themes are post-market surveillance and lifecycle management, so it suits products already close to real-world use. And if you are simply curious, the Commission's blueprint and the government response are both public documents. You do not need permission to read what is being decided about your care.
What to know
On 6 October 2026 the government accepted all 44 recommendations of the National Commission into the Regulation of AI in Healthcare, which reported on 10 September 2026. The central conclusion is that AI medical devices should be monitored throughout their working life rather than assessed once. The MHRA opened applications for AI Airlock Phase 3, focused on post-market surveillance, backed by three years of additional funding, with an applicant webinar on 22 October. Draft guidance on managing changes to AI devices is promised by December 2026, a consultation on qualifying and classifying AI devices is due next year, and staged authorisation pathways are being explored. A full implementation roadmap across all 44 recommendations is due by spring 2027. Sources: Medicines and Healthcare products Regulatory Agency, "Government backs recommendations of NHS doctors-led AI Commission", 6 October 2026, https://www.gov.uk/government/news/government-backs-recommendations-of-nhs-doctors-led-ai-commission ; Medicines and Healthcare products Regulatory Agency, "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 ; Medicines and Healthcare products Regulatory Agency, "AI Airlock: the regulatory sandbox for AIaMD", accessed 8 October 2026, https://www.gov.uk/government/collections/ai-airlock-the-regulatory-sandbox-for-aiamd This article is for general information and does not replace advice from a doctor, pharmacist or other qualified healthcare professional.
Related articles
Four months early: 3,880 rural premises and an £8.2 million contract
On 13 October 2026, the government announced that a rural broadband project in North Oxfordshire had finished four months ahead of schedule. The work connected…
First hearing from a police cell: £17.4 million and four force areas
On 12 October 2026, the Ministry of Justice published the government's response to Lord Leveson's review of the criminal courts. The review made 180 recommendat…
Over 1,000 weak spots found, and a minister's message to AI firms
On Monday 12 October 2026, the UK government set out a package of measures aimed at protecting the country from increasingly powerful artificial intelligence.…