The UK medicines and devices regulator has launched a new programme in Manchester to trial promising healthcare technologies in live NHS environments, seeking earlier evidence of safety, effectiveness and system impact before wider adoption.
Real-world testing with regulator oversight
The Medicines and Healthcare products Regulatory Agency (MHRA), in partnership with Manchester University NHS Foundation Trust (MFT), has established the Manchester Sandbox to bring together developers, clinicians and regulators to assess how new tools perform in routine clinical practice. The initiative is intended to provide clearer evidence about how innovations work outside controlled trials and to identify implementation barriers at an early stage.
According to the announcement, the sandbox will initially concentrate on technologies such as AI-enabled medical devices that aim to identify patients at increased risk of complications from long-term conditions and to support earlier clinical intervention. The programme will deploy technologies in NHS settings with regulatory oversight so that the MHRA and NHS partners can examine safety, effectiveness and the effects on clinical workflows.
“Manchester is leading the way in getting the best new technologies into the hands of local patients faster,” said James Frith, Health Innovation Minister, in a press release. “But innovation must never come at the expense of safety.”
What the sandbox aims to do
The Manchester Sandbox addresses a persistent challenge for healthcare innovators: demonstrating that a device or software performs safely and usefully in day-to-day clinical care, not only in development or controlled studies. For developers, the programme could provide earlier clarity on evidence requirements for broader NHS use and reveal regulatory or implementation hurdles prior to scale-up.
- Assess safety and effectiveness of new technologies in routine care settings.
- Observe impacts on clinical workflows and service delivery.
- Give developers guidance on evidence expectations for NHS adoption.
The MHRA has framed the initiative as a complement to earlier work on artificial intelligence regulation. It follows the agency’s AI Airlock programme, established in 2024 to explore regulatory issues for AI as a medical device. The MHRA also secured £3.6 million in funding earlier this year to extend that work through 2029; the Manchester Sandbox represents a further step toward testing technologies alongside NHS partners before wider deployment.
| Programme | Notable detail |
|---|---|
| AI Airlock | Launched 2024; funding of £3.6m to 2029 |
| Manchester Sandbox | New MHRA–MFT collaboration to test technologies in NHS settings |
Context and limitations
Testing innovations in real-world clinical settings can reveal problems that do not appear in development or controlled trials: integration issues with electronic health records, workflow disruption, unintended biases in AI tools and differences in performance across patient groups. The sandbox model is designed to surface such issues while maintaining regulatory guardrails, but it does not remove the need for robust evaluation or long-term monitoring.
While the announcement highlights the potential for faster local access to beneficial technologies, the evidence generated in Manchester will still need careful interpretation. Results observed in one trust may not generalise across the varied organisational structures, patient populations and IT systems that make up the NHS. Furthermore, real-world deployment requires clear plans for data governance, clinical responsibility and patient consent—all aspects that must be addressed alongside technical performance.
Regulators and health services internationally are increasingly experimenting with sandboxes and pilot environments to balance innovation with safety. The Manchester Sandbox will be watched closely by other NHS organisations, device manufacturers and clinicians for lessons on how to scale promising technologies without compromising patient safety.
The MHRA and MFT have not published a detailed timetable for specific deployments or an exhaustive list of technologies to be included; the announcement indicates the first projects are expected to involve AI-enabled devices supporting the management of long-term conditions.
As the programme develops, peer-reviewed evaluation of outcomes from sandbox trials will be important to distinguish correlation from causation in any observed benefits and to document unintended harms. Transparent reporting and independent scrutiny will be essential if the sandbox is to inform national decisions about regulatory standards and NHS adoption pathways.
