In late June, UpDoc announced that the FDA had cleared its diabetes app in December, which the company bills as the first FDA-cleared medical software using “patient-facing large language models.” Patients interact by voice or text, and the app guides them through insulin adjustments under a doctor-defined treatment plan. The dosing logic is a locked-down, deterministic algorithm, while the language model handles conversation; the FDA cleared the app in the same product category as drug dose calculators. The FDA has authorized more than 1,200 AI-enabled medical devices, none for mental health. Harvard Business Review’s 2025 ranking listed “therapy/companionship” as the number one use of generative AI, and OpenAI disclosed that over a million people a week talk to ChatGPT about suicide. A dozen product-liability cases over ChatGPT’s role in suicides and mental-health harms are consolidated in San Francisco Superior Court. Illinois banned AI-provided therapy in August 2025, Rhode Island followed in June, and the FTC has compulsory orders to seven companies whose chatbots act as companions. In May, Pennsylvania sued Character.AI for unauthorized practice of medicine, alleging its bots posed as licensed psychiatrists.