Guides › UK AI rules and policy

UK AI rules and policy: what it means for health tech start-ups

How is AI in healthcare regulated in the UK?

Updated 6 November 2026 from public records. 16 items below.

At a glance

The UK has no single AI law. Regulators apply the rules they already have, on data, consumer protection, medicines, financial services and more, to AI products in their areas, while government sets out plans on compute, skills and the most powerful models. For a start-up, what matters is which regulator's existing rules already reach its product, and what each is saying about AI this year. AI on GOV.UK.

What it means for your start-up

An AI product that diagnoses, predicts or guides treatment is usually software as a medical device, regulated by the MHRA, which is reworking its rules for AI and runs the AI Airlock sandbox. Selling into the NHS also brings its own checks, such as the Digital Technology Assessment Criteria, and the data protection rules on health data apply throughout.

What to do now

  1. Decide early whether your product is a medical device. The answer sets your route to market and your timeline.
  2. Follow the MHRA's software and AI programme and its sandbox calls, which are open to small firms.
  3. Prepare the NHS assurance evidence (clinical safety, data protection, technical security) before your first pilot asks for it.

Latest from government

Marked: mentions health tech directly.

In Parliament

Marked: mentions health tech directly.

Want this for your own company? Seam tracks what government, Parliament and the regulators are doing and tells you only what affects your business, as it happens.

Ask for a free week

Related guides