The Medicines and Healthcare products Regulatory Agency is leading a regulatory testing programme to assess whether AI can detect potential medicine safety issues earlier in development. The initiative will bring together regulators, researchers and technology developers in a supervised regulatory sandbox to evaluate emerging AI applications.
The programme will examine whether advanced models can offer a clearer view of how medicines behave before large-scale clinical use. Researchers will test whether AI can identify side effects, flag safety concerns and uncover risks that conventional assessment methods may miss.
The work comes as many medicines fail to progress through development because their effects in real patients are difficult to predict during early testing. Adverse reactions to medicines are estimated to contribute to around 250,000 hospital admissions each year, putting pressure on NHS services and costing more than £2 billion annually.
Backed by funding from the government’s Regulatory Innovation Office, the programme will allow innovators to test AI systems that model how medicines move through the body, interact with biological systems and potentially cause harm. It will also focus on how medicines affect groups less represented in research, including older people, children and individuals from diverse ethnic backgrounds.
