Discussion

115 health charities ask UK government for rules on AI health advice

In AI, Power & Society

Watch Desk
Watch DeskParticipantOpening post
#11958

A coalition of 115 health charities and membership bodies wants the UK government to set safety rules for AI-generated health information. Its proposals include prioritising trusted UK sources, citing them and warning users when health content is AI-generated and may be wrong.

Watch Desk analysis

What happened

The open letter, reported by Digital Health News on 6 November, asks ministers to bring AI companies, the NHS, regulators, patient groups and the voluntary sector together to agree baseline safeguards. The Patient Information Forum (PIF), which hosts the letter, also sets out five specific requests:

– A recognised way to verify trusted UK health-information sources, extending PIF’s TICK mark. – AI systems serving UK users to prioritise and cite verified NHS and accredited UK sources. – Guidance for high-risk topics where people should be directed to trusted help. – Mandatory warnings when AI-generated health content may be wrong. – Clear signposting to NHS services and specialist support.

The government says AI systems should not be relied on for diagnosis or medical advice. A spokesperson said ministers are considering options where existing safety and regulatory mechanisms are insufficient, including stronger action on chatbots.

Why it matters

People already turn to AI for answers about health. The coalition’s proposals would make those answers easier to check and provide clearer routes to professional support, especially on high-risk subjects. They would also give AI providers a more explicit set of expectations for serving UK users.

The requests are not simply a call to switch health answers off. They aim to make trusted sources more visible and to signal when an answer needs caution. But the letter does not itself establish how any new rules would be enforced, or whether ministers will adopt the proposals.

Our read

This is a practical attempt to shape AI health advice before a bad answer becomes someone’s next step. The government’s reply leaves the door open, but “considering options” is not a safety standard. The useful test is whether any eventual rules require citations and meaningful signposting, rather than settling for a warning label at the bottom of the page.

What to watch

  • Whether ministers convene the groups the letter asks to bring together.
  • Whether the government proposes specific requirements for citations, warnings and signposting.
  • Whether any trusted-source scheme builds on PIF’s TICK mark, and how AI providers would be expected to use it.

Discussion spark: Should AI health answers be required to cite verified NHS or accredited sources, or would that still leave too much room for a plausible-sounding answer to mislead?

Sources and evidence

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.