Software — including artificial intelligence (AI) — plays an essential part in health and social care. In the United Kingdom, many of these products are regulated as medical devices (or as in vitro diagnostic medical devices, IVDs). The Medicines and Healthcare products Regulatory Agency (MHRA) is the UK regulator responsible for ensuring these products are acceptably safe and meet clinical needs.
For Hong Kong healthcare professionals — whether in Hospital Authority settings, private practice, radiology, pathology, pharmacy, or allied health — understanding how a major regulator frames software as a medical device (SaMD) and AI as a medical device (AIaMD) helps when evaluating products marketed for the UK, products seeking international harmonisation, or tools whose documentation references MHRA guidance.
Software Group: MHRA's dedicated SaMD team
Software Group is responsible for taking all reasonable steps to assure the safety of SaMD and ensuring the UK public have access to technology that meets a clinical need. The group works across the MHRA to achieve this aim for SaMD and AIaMD through:
- assisting with pre-market and post-market enquiries from manufacturers
- conducting technical file reviews and post-market surveillance activities
- reviewing technical and clinical aspects of clinical investigations and exceptional use authorisations
- ensuring medical device regulation is fit for purpose, meets the needs of software as well as AI, and is supported by robust guidance
- engaging with stakeholders including industry, healthcare organisations and professionals, as well as patients and the public
SaMD and AIaMD: key terms
- SaMD (software as a medical device) — standalone software that meets the definition of a medical device, including IVD software.
- AIaMD (AI as a medical device) — SaMD that incorporates artificial intelligence or machine learning, posing additional challenges around transparency, adaptivity, and evidence.
MHRA guidance explains when software applications are considered medical devices and how they are regulated. As Software Group progresses its change programme (described in the next article), existing guidance will be updated as required.
UK regulatory framework in context
MHRA announced an extensive change programme and supporting roadmap to drive regulatory reforms across the SaMD lifecycle — from qualification and classification to requirements that apply both pre- and post-market. The programme also addresses challenges posed by AIaMD, including:
- ensuring these devices are appropriately evidenced
- transparency of AI (explainability and interpretability)
- adaptivity (retraining of AI models)
International harmonisation
Software Group collaborates with international regulators and standards bodies, including:
- the International Medical Device Regulators Forum (IMDRF) Software as a Medical Device Working Group
- the IMDRF Artificial Intelligence/Machine Learning-enabled Working Group
MHRA, the U.S. FDA, and Health Canada have jointly published Good Machine Learning Practice (GMLP) guiding principles, predetermined change control plan (PCCP) principles, and transparency principles — all covered in subsequent articles in this learning path. Hong Kong clinicians evaluating AI tools may encounter these frameworks in vendor dossiers regardless of whether a product holds UK CE/UKCA marking.
Why Hong Kong professionals should care
Hong Kong operates under the Medical Device Administrative Control System (MDACS) administered by the Department of Health, alongside data protection under the Personal Data (Privacy) Ordinance (PDPO). While local rules differ from UK law, MHRA's work on SaMD and AIaMD:
- shapes international consensus through IMDRF and trilateral FDA/Health Canada collaboration
- provides a shared vocabulary for intended use, vigilance, transparency, and change management
- informs how UK-origin products document safety and effectiveness before entering Asian markets
When reviewing an AI-enabled diagnostic or decision-support tool, ask whether the manufacturer can point to qualification guidance, intended purpose documentation, post-market vigilance processes, and transparency materials aligned with frameworks such as those MHRA publishes.
What this learning path covers
The following articles explore the MHRA change programme roadmap, the five PCCP guiding principles, and the transparency framework for machine learning-enabled medical devices — each with practical implications for Hong Kong healthcare settings.
Source: MHRA — Software and artificial intelligence (AI) as a medical device (Updated 3 February 2025)