MHRA launches AI sandbox for medicines development and safety testing
The UK's Medicines and Healthcare products Regulatory Agency (MHRA) will launch a new AI sandbox this summer, giving companies and researchers a controlled space to test AI systems for predicting drug safety and side effects.
UK Health Innovation Minister Preet Gill announced the programme during London Tech Week. The sandbox will allow innovators to work directly with regulators to explore how AI can improve drug safety assessments and forecast risks that current methods miss.
The initiative addresses a concrete problem. Adverse drug reactions cause around 250,000 hospital admissions annually in the UK, costing the NHS over £2 billion. Roughly 90% of medicines fail during development because existing methods cannot reliably predict how drugs will perform in real patients.
How the sandbox will work
The initial phase will include up to five AI-driven approaches. The MHRA will collaborate with industry and academic groups from summer 2026 to define operational details and success criteria.
Participating teams will test AI systems to assess how medicines behave in the body, improve predictions about drug absorption and processing, and identify potential harms. The sandbox will also explore how AI can incorporate clinical data from underrepresented groups-children, older adults, and people from diverse ethnic backgrounds.
The UK Government's Regulatory Innovation Office is funding the programme.
What comes next
The MHRA will use findings from the sandbox to evaluate whether AI tools can reliably inform decisions about new medicines' safety. Clear evidence and defined expectations for AI use in drug development should encourage investment in UK healthcare innovation.
The sandbox forms part of broader efforts to modernise medicine development through advanced modelling and synthetic data. In March 2026, the MHRA and the National Institute for Health and Care Excellence began reviewing new drugs concurrently, following changes to the UK's regulatory process.
For IT and development professionals building AI systems for regulated industries, understanding how regulators assess AI reliability is increasingly critical. Learn more about AI implementation for IT and development teams.
Your membership also unlocks:
