An AI-enhanced NHS will rely on clinical guidelines, but what if the evidence is inequitable?

The UK’s new 10-Year Health Plan for England sets out an ambitious agenda for an increasingly digitised and data-driven healthcare system, where artificial intelligence (AI) is expected to play a major role in realising long-term aims to deliver personalised medicine for all. A recent study highlighted poor consideration of sex and gender in National Institute for Health and Care Excellence clinical guidelines, which will be used to train AI models used by the NHS. When AI models are trained on inequitable data, they risk amplifying bias, with harmful implications for patient safety.