Overview
- - A practitioner guide details five immediate steps for providers: clarify website intake language, add AI-aware triage scripts, treat AI-generated notes as patient-supplied information, place targeted privacy warnings, and set cross‑functional protocols.
- - A Jan. 21 survey of more than 20,000 U.S. adults found 10.3% use generative AI daily, and 87.1% of those users turn to it for personal advice or emotional support.
- - The American Psychological Association advises against using chatbots as a substitute for therapy or crisis care, with 988 remaining the recommended resource in emergencies.
- - Investigative reporting has documented serious harms when crises were mishandled, including a New York Times tally of nearly 50 crisis incidents during ChatGPT conversations and three deaths.
- - OpenAI, Anthropic and Google say they are working with clinicians to improve responses to sensitive conversations, as experts note both practical benefits like thought‑organizing and risks including privacy exposure and reinforcement of unhealthy patterns.