Can an NHS Trust Blame the Developer When Diagnostic AI Fails? What the UK Jurisdiction Taskforce's July 2026 Liability Statement and the MHRA Commission's Recommendations Say

What the UK Jurisdiction Taskforce concluded in July 2026

The UK Jurisdiction Taskforce, chaired by the Master of the Rolls, published its 130-page Legal Statement on Liability for AI Harms in July 2026 after a public consultation on a draft. Its central finding is that the existing private law of England and Wales can resolve claims for harm caused by AI without a new liability regime. AI has no legal personality, so liability attaches to the developers, providers and users behind it. Contract allocates risk along an AI supply chain first, and negligence fills the gaps.

Why hospitals and clinicians read it closely

The statement treats an NHS Trust's duty to its patients as non-delegable: a Trust cannot escape liability for a defective diagnostic tool by pointing at the developer. It also says a clinician may be liable for failing to use AI where a competent practitioner would have used it; its own example is a radiologist who ignores an effective tumour-detection tool available at reasonable cost. Browne Jacobson's 30 October 2026 reading sets the statement beside the MHRA National Commission's recommendations of 10 September 2026. This post asks who answers when a diagnostic AI fails, and where the EU's new Product Liability Directive draws the line differently by treating software as a product.