Software and AI as medical devices (SaMD/AIaMD)

Software as a Medical Device (SaMD) is standalone software that qualifies as a medical device in its own right. It is not software that is part of a hardware medical device (which is regulated as part of the hardware device).

SaMD is regulated under the UK Medical Devices Regulations 2002 and must be registered with MHRA before being placed on the GB market. AI as a Medical Device (AIaMD) is a subset of SaMD that uses artificial intelligence or machine learning algorithms.

Is your software a medical device?

Software is a medical device if it has a medical intended purpose, such as:

Diagnosis, prevention, monitoring, prediction, prognosis, or treatment of disease
Diagnosis, monitoring, treatment, or alleviation of an injury or disability
Investigation, replacement, or modification of anatomy or a physiological process

Software that only stores, archives, communicates, or performs simple searches of patient data without processing or analysing it for a medical purpose is not a medical device.

MHRA AI medical device requirements

Requirements for AI as a Medical Device (AIaMD) and Software as a Medical Device (SaMD) under UK medical device regulations.

AI as a Medical Device (AIaMD) and Software as a Medical Device (SaMD) must be registered with MHRA before being placed on the UK market. The MHRA has a dedicated Software Group developing regulatory requirements for AI-powered medical devices. Manufacturers must apply Good Machine Learning Practice (GMLP) principles.

Registration required: All AIaMD and SaMD must be registered with MHRA before being placed on the UK market
Current classification: SaMD classified under Medical Devices Regulations 2002 Annex IX — currently can be Class I to III depending on intended purpose and risk
Future reclassification (expected 2026): MHRA will adopt IMDRF SaMD risk framework — most SaMD will be reclassified to Class IIa minimum, requiring UK Approved Body involvement (up-classification for many currently Class I products)
International reliance: Streamlined approval for devices already authorised by FDA (US), Health Canada, TGA (Australia), or EU Notified Bodies — excludes SaMD
CE mark transition: CE marked SaMDs recognised in UK until 30 June 2028 (old Directives) or 30 June 2030 (EU MDR certified)
Good Machine Learning Practice (GMLP): MHRA expects AIaMD manufacturers to apply GMLP principles — transparency in algorithm design, appropriate clinical evidence, predetermined change control plans for adaptive algorithms
Post-market surveillance: Enhanced PMS requirements from 16 June 2025 — manufacturers must monitor via MORE portal, report incidents within 2/10/15 day statutory timelines
Criminal offence: Placing a non-compliant AI medical device on the UK market — up to 6 months imprisonment and/or unlimited fine

Current classification

Under the current UK MDR 2002, SaMD is classified using the same Annex IX criteria as other general medical devices. Many SaMD products are currently classified as Class I, allowing self-certification without UK Approved Body involvement.

Future reclassification

MHRA will adopt the IMDRF SaMD risk categorisation framework, which considers:

The significance of the information provided by the SaMD to the healthcare decision
The state of the healthcare situation or condition

Under the new rules, most SaMD will be reclassified as Class IIa at minimum, requiring UK Approved Body involvement. This represents an up-classification for many currently Class I SaMD products. The reclassification is expected via a statutory instrument in 2026.

Good Machine Learning Practice (GMLP)

MHRA expects AIaMD manufacturers to apply GMLP principles:

Transparency: Document algorithm design, training data sources, and known limitations
Clinical evidence: Demonstrate clinical validity and safety appropriate to the risk classification
Predetermined change control plans: For adaptive algorithms that learn over time, establish predetermined change control plans describing what changes the algorithm may make without requiring a new conformity assessment
Monitoring: Implement robust post-market surveillance for algorithm performance drift

Medical device classification system

Risk-based classification system for medical devices determining regulatory pathway.

Class I (lowest risk): Non-invasive devices with minimal risk — self-certification by manufacturer (e.g., bandages, wheelchairs, stethoscopes)
Class Is (sterile, low risk): Sterile low-risk devices — self-certification plus UK Approved Body assessment for sterility aspects (e.g., sterile bandages, sterile surgical gloves)
Class Im (measuring, low risk): Low-risk devices with measuring function — self-certification plus UK Approved Body for metrological aspects (e.g., non-invasive thermometers, graduated syringes)
Class IIa (low-medium risk): Short-term invasive devices and diagnostic equipment — UK Approved Body assessment required (e.g., hearing aids, dental crowns, blood pressure monitors)
Class IIb (medium-high risk): Long-term invasive and surgically invasive devices — UK Approved Body full assessment required (e.g., ventilators, blood bags, condoms)
Class III (highest risk): Devices contacting heart or CNS, containing medicinal substances, or made from animal tissue — UK Approved Body full design examination and QMS assessment (e.g., heart valves, hip implants, coronary stents, pacemakers)
Active implantable medical devices (AIMD): Devices totally or partially introduced into the human body and intended to remain — UK Approved Body full assessment required (e.g., cardiac pacemakers, cochlear implants, neurostimulators)
Classification basis: Classification determined by intended purpose, invasiveness, duration of body contact, and whether device is active — rules in Annex IX of retained Directive 93/42/EEC

Medical device classification guidance

Business Context Testing

Content Graph

Guvnor

Software and AI as medical devices (SaMD/AIaMD)

Is your software a medical device?

Current classification

Future reclassification

Good Machine Learning Practice (GMLP)

Compliance Assistant

🧪 Business Context Testing

◉ Content Graph

Is your software a medical device?

Current classification

Future reclassification

Good Machine Learning Practice (GMLP)

Related guides in General Compliance

Business Context Testing

Content Graph