The UK's Medicines and Healthcare Products Regulatory Agency (MHRA), responsible for overseeing medical devices and drug approvals, has proposed new guidelines for AI technologies used in the National Health Service (NHS) and other healthcare environments. These recommendations, developed by an independent commission that gathered input from over 12,000 individuals—including patients, doctors, and healthcare professionals—emphasize the need for ongoing monitoring of AI systems. If an AI product malfunctions or becomes less effective, it could be removed from regulatory approval. Patients would also be informed if AI is involved in their care and given access to details about the specific AI tools being used. The report also suggests imposing penalties on AI developers who fail to meet safety and effectiveness standards, and introduces an "AI L plate" system, which would allow new AI models to be tested by healthcare professionals under strict supervision, similar to how new drivers are monitored with a learner’s permit. MHRA chief Lawrence Tallon told the BBC that AI will soon become a regular part of NHS operations, with patients increasingly seeing it as a standard component of their care. He noted that the existing framework for regulating medical devices, which was created for products like hip replacements and stethoscopes, is not suitable for AI, which can continuously evolve after being approved. Tallon acknowledged that regulating AI is a challenge facing all countries, as no single regulatory system has yet fully addressed the complexities of AI in healthcare. AI-powered scribes, which use large language models (LLMs) to take notes and generate reports during patient consultations, are being used by about 40% of general practitioners (GPs) in the UK. However, a recent study from the University of Edinburgh found that patients may be less willing to share sensitive personal information, such as substance abuse history, if they know their conversations are being processed by AI. Professor Henrietta Hughes, a GP and member of the report commission, said while some patients are comfortable with AI during consultations, others prefer to avoid it, saying, "I don't want to talk to a robot." She noted that AI scribes can make errors in their notes but emphasized that doctors are responsible for reviewing and correcting them. Currently, AI is more commonly used for administrative tasks or to assist in identifying symptoms under the supervision of human experts. However, there is growing excitement about its potential in health and medical research. Professor Alastair Denniston, an ophthalmologist involved in the report, called AI "an exceptional opportunity" for healthcare, comparing its potential impact to major medical breakthroughs like antibiotics and MRI technology. Meanwhile, chip designer Arm's CEO, Rene Haas, has predicted that AI could lead to a cancer cure within our lifetime. Despite these optimistic views, concerns remain about AI systems making biased or incorrect decisions based on flawed training data and the risk of AI chatbots providing inaccurate medical advice.