MHRA secures £3.6m to expand AI regulatory sandbox

The Medicines and Healthcare products Regulatory Agency (MHRA) has announced a £3.6 million funding boost to scale its AI Airlock programme over the next three years, signalling a significant expansion of the UK’s regulatory infrastructure for artificial intelligence in healthcare.

The investment will support additional cohorts of AI developers and broaden the scope of the Airlock, a regulatory sandbox designed to test AI-enabled medical devices in controlled, real-world settings. The programme aims to help regulators and suppliers better understand how to evaluate safety, performance and compliance for emerging AI technologies.

Launched in 2024, the AI Airlock brings together regulators, the NHS, and industry partners to explore regulatory challenges associated with AI as a medical device (AIaMD). It enables companies to work directly with the regulator to generate evidence and identify potential risks before wider deployment.

The new funding will allow the MHRA to increase the number of technologies entering the sandbox and extend collaboration across the health system. It also supports the development of regulatory guidance informed by real-world testing.

The move builds on earlier phases of the programme, including a pilot and a second cohort running through to 2026, which have tested a range of applications such as AI-assisted diagnostics and clinical decision support tools.

The sandbox model is intended to address gaps in existing medical device regulation, particularly where AI systems evolve over time or rely on complex data inputs. Insights from the programme are expected to feed into future policy and standards for AI-enabled healthcare products.

However, the programme does not provide a direct route to market approval. Participating organisations are expected to generate their own evidence and fund development activities, with the Airlock focusing on regulatory insight rather than certification.