AI therapy chatbots surge
- AI therapy chatbots are proliferating rapidly but many lack robust clinical evidence and clear regulation. - Social posts linking KFF Health News have highlighted the surge and warned about safety and effectiveness uncertainties. (x.com) - Unregulated chatbots increase liability and effectiveness uncertainty when integrated into brief behavioral‑health workflows. (x.com)
Artificial intelligence therapy apps are spreading faster than the rules and research meant to test whether they are safe. (kffhealthnews.org) KFF Health News reported on April 17 that it identified about 45 artificial-intelligence therapy apps in Apple’s App Store in March, marketed for anxiety, depression and other mental-health needs. (kffhealthnews.org) The pitch lands in a strained system: KFF Health News reported in earlier coverage that more than half of U.S. counties lack psychiatrists, creating an opening for cheaper, always-on chatbot products. (kffhealthnews.org) These products often sit in a gray zone between wellness app and medical device. The Food and Drug Administration said in 2025 that its Digital Health Advisory Committee was examining how to regulate artificial-intelligence tools used in therapy and other mental-health care. (fda.gov) Food and Drug Administration briefing materials for that meeting said the evidence standards used for prescription digital therapeutics and diagnostics may also apply to generative-artificial-intelligence mental-health devices, including postmarket monitoring after release. (fda.gov) Outside that regulated slice, many apps still reach consumers without independent proof they improve symptoms. The American Psychiatric Association says clinicians and patients should review privacy, evidence, usability and interoperability before relying on a mental-health app. (psychiatry.org) States have started to move ahead of Washington. Utah enacted House Bill 452 in March 2025 to require disclosures and privacy protections for mental-health chatbots used by people in the state. (le.utah.gov) Illinois went further in August 2025, when Gov. J.B. Pritzker signed House Bill 1806 barring licensed professionals from using artificial-intelligence systems to make independent therapeutic decisions or directly communicate with clients as therapy without clinician review. (ilga.gov) The safety debate has sharpened around young users. RAND wrote in September 2025 that a Common Sense Media survey found 72% of U.S. teenagers had used artificial-intelligence chatbots as companions, and nearly one in eight had sought emotional or mental-health support from them. (rand.org) Common Sense Media said in November 2025 that testing with Stanford Medicine’s Brainstorm Lab found major chatbots remained unsafe for teen mental-health support, even after some improvements in responses to suicide and self-harm prompts. (commonsensemedia.org) The result is a market where demand, access gaps and software speed are moving ahead of clinical proof. Regulators, clinicians and app makers are now arguing over whether these bots should be treated more like self-help tools, medical products or something in between. (kffhealthnews.org)
