What changed
MHRA updated its “Software and artificial intelligence (AI) as a medical device” guidance page on 3 February 2025 and, within it, references joint MHRA–FDA–Health Canada guiding principles for (1) PCCPs for machine-learning devices and (2) transparency for machine-learning medical devices. MHRA also continues to signpost its Software and AI as a Medical Device change programme roadmap (updated 14 June 2023), which lays out work packages spanning qualification, classification, premarket requirements, post-market, cybersecurity, interpretability and adaptivity.
Why it matters
For SaMD/AIaMD—especially adaptive ML medical devices—post-deployment changes can affect performance and therefore influence post-market surveillance and medical device vigilance obligations. MHRA’s emphasis on PCCP and transparency principles (aligned with FDA and Health Canada) signals that UK-facing compliance and inspection-readiness may depend on how well organizations can demonstrate controlled, pre-planned change governance and clear transparency artifacts, alongside post-market/vigilance processes that account for algorithm updates.
Practical implication
Ask the safety/device vigilance and PV CSV leads to: (1) confirm whether any marketed/fielded SaMD/AIaMD uses adaptive ML or planned updates that would fit a PCCP-like approach; (2) review current change-control SOPs and technical documentation to ensure change governance and transparency artifacts are captured and retrievable; (3) verify post-market/vigilance processes explicitly account for algorithm updates (e.g., monitoring after updates, escalation criteria).