Home / Technology / UK Healthcare AI Needs New Rules, Watchdog Says
UK Healthcare AI Needs New Rules, Watchdog Says
10 Sep
Summary
- UK watchdog proposes 44 recommendations for AI in healthcare.
- Patients must know when AI is used in their care.
- AI products must be continuously monitored and removable if faulty.
The UK's Medicines and Healthcare Products Regulatory Agency (MHRA), the nation's medical device regulator, has outlined 44 recommendations for updated policies concerning Artificial Intelligence in healthcare. MHRA chief Lawrence Tallon anticipates AI will soon be a standard component of NHS healthcare delivery, stressing the importance of maintaining patient confidence.
These recommendations, developed with input from over 12,000 individuals including patients and clinicians, address the dynamic nature of AI. They propose continuous monitoring of AI products, ensuring they can be removed if they malfunction or lose effectiveness. Furthermore, patients should be informed about AI's involvement in their treatment and have access to product details.
The MHRA also advocates for penalizing AI developers who fail to meet standards and introducing an "AI L plate" system for supervised trials of new AI models. Current regulations, designed for static medical devices, are insufficient for AI that evolves post-authorization as it learns from new data. This presents a global challenge, as no single country has fully resolved AI regulation.
AI scribes are already used by 40% of UK GPs, though patient willingness to share sensitive information with AI varies. While acknowledging potential AI errors in note-taking, doctors are responsible for corrections. The potential of AI in health and medical research is significant, comparable to breakthroughs like antibiotics or MRI scans.
