UK Launches New MHRA Regulatory Sandbox for AI-Enabled Medical Devices: What London Region I Means for Regulatory Affairs

4 days ago
13 min read

The MHRA AI regulatory sandbox for medical devices, known as London Region I, allows up to 10 AI-enabled medical device manufacturers to deploy their technologies in real NHS clinical settings under regulatory oversight while generating real-world safety and effectiveness evidence. This affects AI medical device manufacturers, NHS providers across London, and UK regulatory affairs professionals working in medical devices. It applies from the Expression of Interest stage opened on 30 July 2026, with the programme running until December 2028, and it matters now because it signals how the MHRA intends to bridge the gap between regulatory approval and real-world NHS adoption of AI technologies.

MHRA London Region I AI regulatory sandbox — overview for AI-enabled medical device manufacturers

1. At a Glance

2. Background: Why the MHRA's AI Sandbox Approach Was Needed

3. Inside London Region I: How the Sandbox Actually Works

4. How London Region I Relates to the MHRA's Earlier AI Airlock Programme

5. The Wider UK Medical Device Reform Context

6. Related Regulatory Affairs Training

7. Practical Implications for Regulatory Affairs Professionals

8. Key Takeaways

9. Take the Next Step in Your Regulatory Affairs Career

10. Frequently Asked Questions

11. Further Reading and Reference Sources

At a Glance

- The Medicines and Healthcare products Regulatory Agency (MHRA) launched London Region I, a new regulatory sandbox for AI-enabled medical devices, first published on 10 June 2026.

- Expressions of Interest opened on 30 July 2026 for both AI medical device manufacturers and NHS providers.

- Up to 10 AI medical device manufacturers will be selected for the initial phase, working alongside NHS providers across London.

- The programme is delivered in partnership with NHS England (London) and three London Health Innovation Networks (HINs): Imperial College Health Partners, UCLPartners, and HIN South London.

- The sandbox is expected to run until December 2028, generating real-world safety and effectiveness evidence under MHRA oversight.

- London Region I is distinct from the MHRA's earlier AI Airlock programme, whose Phase 2 report (covering seven AI technologies) was published on 9 June 2026.

- The sandbox launch coincides with the MHRA's draft Medical Devices (Amendment) Regulations 2026, which propose new pathways for AI and software medical devices.

Background: Why the MHRA's AI Sandbox Approach Was Needed

Artificial intelligence-enabled medical devices present a distinctive regulatory challenge: many are designed to learn, adapt, or be updated after they reach the market, which sits uncomfortably alongside a traditional medical device regulatory model built around a fixed product receiving a single pre-market approval. The MHRA has been working since 2024 to address this gap, and by mid-2026 the pressure to move from pilot testing to real-world deployment had become acute. NHS organisations across the UK are under sustained pressure to adopt digital and AI tools that can expand access to care and reduce health inequalities, as set out in the government's 10 Year Health Plan, but manufacturers and NHS trusts alike have lacked a clear, safety-assured route for testing AI medical devices directly within live clinical environments before wider procurement decisions are made.

Interested in building a career in regulatory affairs and working with frameworks like the MHRA's regulatory sandboxes for AI-enabled medical devices? The FULL Introduction to Regulatory Affairs Course at Entry to Regulatory covers UK medical device regulation, software and AI as a medical device classification, and post-market surveillance requirements, giving you the practical knowledge, real work experience and job search support to break into or advance within regulatory affairs. - 50+ CPD-accredited hours of expert-led training - Up to 3 months of real regulatory work experience - Mentoring and job search support until you are employed - Covers EU, UK and US pharmaceutical regulations - Taught by a regulatory affairs expert with 10+ years at MHRA, GSK, MSD and Bayer - Study online, part-time, just 6 hours per week View the course

The regulatory groundwork for this moment stretches back further than most coverage suggests. The MHRA launched its first dedicated AI regulatory sandbox, known as the AI Airlock, on 9 May 2024, specifically to explore the regulatory challenges of standalone AI as a Medical Device (AIaMD), with applications opening on 23 September 2024. That programme's second phase ran from April 2025 to May 2026 and worked with seven AI technologies covering use cases such as clinical note-taking, cancer diagnostics, and obesity management, generating insights on lifecycle management, human oversight, and post-market change control that fed directly into the design of London Region I. Understanding how pre-market sandbox testing differs from real-world post-authorisation deployment sandboxes is a core component of the Entry to Regulatory regulatory affairs training programme, which covers medical device classification and lifecycle regulatory strategy as part of its practical curriculum for EU, UK and US regulations — see the full course details at https://pages.entrytoregulatory.com/courses/

Timeline showing the MHRA's AI Airlock sandbox and London Region I AI medical device sandbox development

Inside London Region I: How the Sandbox Actually Works

London Region I is a real-world regulatory sandbox designed to let AI-enabled medical devices operate in live NHS clinical settings while regulators, providers and manufacturers jointly generate evidence of safety and effectiveness. First published by the MHRA on 10 June 2026 and updated on 30 July 2026 to open the formal call for Expressions of Interest, the programme is delivered jointly by the MHRA, NHS England (London), and three London Health Innovation Networks (HINs) — Imperial College Health Partners, UCLPartners, and HIN South London. Up to 10 AI medical device manufacturers will be selected for the initial phase, working directly alongside NHS providers across London under MHRA oversight, with the programme expected to run until December 2028.

The stated purpose is to close what the organisations involved describe as one of the biggest barriers to healthcare innovation: the gap between formal regulatory approval and actual, trusted adoption within a live health system. Lawrence Tallon, MHRA Chief Executive, said of the launch:

> "This initiative demonstrates that regulation can be an enabler for innovation, not a barrier. We need to work faster to keep pace with developments in AI to ensure that patients can safely benefit from cutting-edge technologies as they become available."

Regulatory professionals looking to deepen their understanding of how real-world evidence generation interacts with post-market surveillance obligations will find relevant practical training in the Entry to Regulatory course, which includes hands-on assignments covering UK medical device regulation and post-approval regulatory strategy as part of its UK regulatory curriculum. Full details at https://pages.entrytoregulatory.com/courses/

How the MHRA's London Region I AI regulatory sandbox connects regulators, NHS providers and manufacturers

How London Region I Relates to the MHRA's Earlier AI Airlock Programme

It is important to distinguish London Region I from the MHRA's earlier and separate AI Airlock sandbox, since both are regulatory sandboxes for AI-enabled medical devices but serve different purposes in a product's lifecycle. AI Airlock was launched on 9 May 2024 as a pre-market regulatory sandbox specifically for standalone AI as a Medical Device (AIaMD), allowing innovators to work through novel regulatory questions — such as adaptive algorithm behaviour and appropriate human oversight — before formal marketing authorisation, in a space that does not carry the legal weight of a real device approval. Applications for AI Airlock opened on 23 September 2024, and its Phase 2 Programme Report, covering 7 case studies including clinical note-taking, cancer diagnostics, and obesity management tools, ran from April 2025 to May 2026 and was published on 9 June 2026 (last updated 27 July 2026).

The Phase 2 report identified several recurring regulatory challenges that manufacturers and the MHRA had to work through together:

1. Lifecycle management — how to regulate a product whose intended purpose may evolve after launch

2. Human oversight — how oversight requirements should adapt as AI models change

3. Change control — how to manage and evidence changing AI behaviour without requiring a new submission for every update

4. Performance evaluation — ensuring performance metrics remain clinically relevant, not just statistically valid

The MHRA has secured funding for a Phase 3 of AI Airlock, intended to translate these insights into actionable regulatory guidance. London Region I, by contrast, operates later in the lifecycle: it assumes devices already have, or are close to, regulatory clearance, and instead focuses on generating real-world evidence during live NHS deployment to build the evidence base needed for wider procurement and adoption decisions. Comparable international activity has moved faster in one specific area — the US Food and Drug Administration (FDA) finalised its own guidance on Predetermined Change Control Plans (PCCPs) for AI-enabled devices on 3 December 2024, giving US manufacturers a defined route to make planned algorithm modifications without resubmitting a full marketing application, a mechanism the UK's AI Airlock work is still developing equivalent guidance for.

The Wider UK Medical Device Reform Context

London Region I does not exist in isolation; it launches alongside a broader legislative overhaul of the UK's medical device framework. The MHRA has published draft Medical Devices (Amendment) Regulations 2026, alongside a stakeholder impact survey, proposing revisions to pre-market requirements for medical devices and in vitro diagnostic devices (IVDs) in Great Britain. Stakeholder feedback on these draft regulations closed on 19 June 2026, with adoption anticipated in December 2026 and implementation expected from June 2027.

Reform Element	What It Covers	Relevance to AI-Enabled Devices
International Reliance Pathway	Streamlined applications for devices already approved by recognised international regulators	Could accelerate UK market access for AI devices cleared by the FDA or other trusted regulators
Revised classification rules	Updated risk-based classification aligned with EU MDR and IMDRF standards	Determines the level of scrutiny applied to AI-enabled software devices
Predetermined Change Control Plans (PCCPs)	New UK pathway for pre-approved device modifications	Directly relevant to AI models that are updated or retrained after launch
Mandatory Unique Device Identifiers (UDI)	Traceability requirements for devices placed on the GB market	Supports post-market surveillance of AI-enabled devices deployed via sandboxes
Expanded technical documentation	Broader evidentiary requirements for higher-risk devices	Increases the documentation burden for AI/software medical device manufacturers

These reforms are explicitly intended to align the UK framework with the current EU Medical Devices Regulation (MDR) and International Medical Device Regulators Forum (IMDRF) standards, while also introducing new pathways tailored specifically to software and AI-enabled devices, including UK-specific PCCP provisions. For regulatory affairs professionals, this means London Region I should be read as one part of a coordinated MHRA strategy — sandbox testing, real-world evidence generation, and legislative reform moving in parallel rather than as isolated initiatives.

Deepen Your Knowledge: Regulatory Affairs Training on This Topic

Working effectively with AI-enabled medical device regulation requires regulatory affairs professionals to understand several interconnected knowledge areas: UK medical device classification rules, the distinction between pre-market and post-market regulatory sandboxes, post-market surveillance obligations, and how the UK framework is diverging from — and realigning with — the EU MDR and US FDA approaches following Brexit. AI-specific regulatory concepts, such as Predetermined Change Control Plans and lifecycle-based risk management, are becoming core professional competencies rather than niche specialisms.

The FULL Introduction to Regulatory Affairs Course at Entry to Regulatory addresses these knowledge areas directly through its coverage of UK medical device regulation, software and AI as a medical device classification, and comparative EU, UK and US regulatory frameworks, supported by real-world case studies drawn from current industry developments such as this one. Rather than treating medical device regulation as a niche add-on, the course integrates it into a broader picture of how regulatory affairs professionals operate across pharmaceuticals, biologics and devices.

This training is most valuable for life science and engineering graduates seeking entry-level regulatory affairs roles in medical devices, healthcare and digital health professionals considering a career change, and scientists or software specialists looking to move into a formal regulatory affairs career. No prior regulatory experience is required, and the practical work experience component gives learners direct exposure to the kind of regulatory frameworks discussed in this article.

TRAINING SPOTLIGHT BOX Course: FULL Introduction to Regulatory Affairs Course Provider: Entry to Regulatory Relevant to this topic: UK medical device regulation, Software and AI as a Medical Device (SaMD/AIaMD) classification, post-market surveillance, EU/UK/US regulatory framework comparison, real-world regulatory case studies Format: Online, part-time | 6 hours per week | 3 months Includes: Up to 3 months real work experience, job mentoring, CV review, mock interviews, industry-recognised certificate Suitable for: Life science graduates, pharmacy professionals, career changers, scientists seeking regulatory transition Start here

Practical Implications for Regulatory Affairs Professionals

AI medical device manufacturers, NHS providers and regulatory affairs professionals are asking practical questions about how London Region I changes their day-to-day regulatory strategy. The table below summarises the most common questions and what has changed.

Key Question	Previous Situation	What Changes Now
How do we generate real-world evidence for an already-authorised AI device?	No dedicated NHS-wide sandbox existed for post-approval real-world deployment evidence	London Region I offers a structured route to deploy in live NHS settings under MHRA oversight from July 2026 EOI stage
Is London Region I the same as applying to AI Airlock?	Manufacturers sometimes conflated pre-market sandbox testing with real-world deployment testing	The two are distinct: AI Airlock addresses pre-market regulatory questions; London Region I addresses post-approval, real-world NHS deployment
How will planned AI model updates be regulated going forward?	No dedicated UK pathway existed for pre-approved AI modifications	The draft Medical Devices (Amendment) Regulations 2026 propose a UK Predetermined Change Control Plan pathway
Can our FDA-cleared AI device get faster UK market access?	UK and US device approvals were assessed independently with no formal reliance mechanism	The proposed International Reliance Pathway may allow streamlined UK applications for devices already cleared by trusted international regulators
What happens if we are not selected for the initial 10 manufacturer places?	No structured queue or alternative route existed	Manufacturers should monitor future MHRA sandbox expansion phases and regional rollouts beyond London Region I
Do we need to change our post-market surveillance approach for AI devices?	Post-market surveillance requirements were largely generic across device types	New PMS regulations effective from June 2025 and lifecycle-based change control concepts from AI Airlock are shaping AI-specific expectations
When should we prepare our impact assessment for the 2026 regulatory reforms?	The Medical Devices Regulations 2002 framework applied with limited AI-specific provision	With adoption anticipated December 2026 and implementation from June 2027, manufacturers should begin gap analysis against the draft regulations now

Key Takeaways

1. Submit an Expression of Interest promptly. If your organisation manufactures AI-enabled medical devices relevant to NHS London priorities, review the eligibility criteria for London Region I without delay, as only 10 places are available in the initial phase.

2. Map your product against the right sandbox. Distinguish clearly between AI Airlock (pre-market regulatory testing) and London Region I (post-approval real-world deployment) before choosing which programme, if either, fits your product's current lifecycle stage.

3. Review your change control documentation now. Begin preparing lifecycle and change control documentation in anticipation of a UK Predetermined Change Control Plan pathway under the 2026 regulatory reforms.

4. Track the Medical Devices (Amendment) Regulations 2026 closely. With adoption anticipated in December 2026 and implementation from June 2027, build a compliance timeline now rather than waiting for the final text.

5. Assess your International Reliance Pathway eligibility. If your device already holds FDA or other trusted-regulator clearance, evaluate whether the proposed reliance route could accelerate UK market access.

6. Engage with your local Health Innovation Network. Build relationships with Imperial College Health Partners, UCLPartners or HIN South London if you plan to deploy AI medical devices within London NHS settings.

7. Benchmark against FDA's PCCP guidance. Use the FDA's finalised December 2024 Predetermined Change Control Plan guidance as a reference point when anticipating the shape of the UK's equivalent framework.

Take the Next Step in Your Regulatory Affairs Career

The launch of London Region I illustrates exactly why regulatory affairs has become such a dynamic career path within medical devices and digital health. Understanding how sandbox programmes, post-market surveillance, and evolving UK legislation like the Medical Devices (Amendment) Regulations 2026 fit together is a core skill for regulatory affairs professionals at every level, whether working directly for a manufacturer, an NHS organisation, or a regulatory consultancy. If this article has sparked your interest in how regulation shapes the adoption of healthcare technology from the inside, the Entry to Regulatory Introduction to Regulatory Affairs course is a practical next step, whether you are completely new to the field or looking to formalise and advance existing experience. A free introductory webinar is available as a low-commitment way to explore whether the course is right for you, with no obligation to enrol.

Explore the full course details and register for a free introductory webinar at Entry to Regulatory: https://pages.entrytoregulatory.com/courses/

Professional studying regulatory affairs training online — medical device regulation career development

Frequently Asked Questions

What is London Region I?

London Region I is a regulatory sandbox launched by the MHRA, in partnership with NHS England (London) and three London Health Innovation Networks, that allows up to 10 AI-enabled medical device manufacturers to deploy their technologies in real NHS clinical settings while generating safety and effectiveness evidence under MHRA oversight. It is expected to run until December 2028.

Is London Region I the same as the MHRA's AI Airlock?

No. AI Airlock, launched in May 2024, is a pre-market regulatory sandbox exploring regulatory questions for standalone AI as a Medical Device before formal authorisation. London Region I focuses on real-world, post-approval deployment within live NHS settings to generate adoption evidence.

How many manufacturers can take part in London Region I?

Up to 10 AI medical device manufacturers will be selected for the initial phase, working directly alongside NHS providers across London.

How does the UK's approach compare to the FDA's regulation of AI-enabled devices?

The FDA finalised guidance on Predetermined Change Control Plans (PCCPs) for AI-enabled devices on 3 December 2024, giving US manufacturers a defined route for pre-approved algorithm modifications. The UK's equivalent PCCP pathway is still being developed through the draft Medical Devices (Amendment) Regulations 2026 and insights from the AI Airlock programme.

How can I start a career in regulatory affairs working on medical device and AI regulation?

A structured, practical training programme is the most direct route into the field, particularly for those without existing regulatory experience. The Entry to Regulatory Introduction to Regulatory Affairs course covers UK medical device regulation, software and AI as a medical device classification, and EU/UK/US regulatory frameworks, alongside job search support and real work experience. Full details are available at https://pages.entrytoregulatory.com/courses/.

Speak To Us