NHS AI Scribes Patient Safety Risks

AI scribes used in NHS consultations are posing patient safety risks by incorrectly transcribing names of drugs, illnesses, and medical diagnoses. Healthwatch England highlighted cases including a patient mistakenly given a traumatic diagnosis of demyelination due to an AI error and transcription mix-ups involving prescription drugs. The government's 10-year health plan heavily relies on AI scribes to reduce bureaucratic burdens and modernize services, but critics warn that the technology causes dangerous hallucinations and errors. Furthermore, the Medicines and Healthcare products Regulatory Agency has chosen not to classify AI scribes as medical devices, leaving a critical lack of national oversight. Medical professionals and patient advocacy groups emphasize that trust and robust communication are currently missing, noting that patients frequently have to catch and correct these AI mistakes themselves. Additionally, the necessity for doctors to meticulously review transcripts diminishes expected time-saving benefits while increasing legal liability risks for the NHS. Experts urge immediate clarity on reporting mechanisms for AI errors and recommend enhanced partnership with patients to safeguard medical records. Without stringent regulatory oversight and improved accuracy, the rapid rollout of ambient voice technology risks compromising the integrity of patient care across England.