US health officials deploy medical AI tools amid safety and oversight concerns

The US government is pushing AI into healthcare faster than the guardrails can keep up. Federal agencies are embracing artificial intelligence tools to address a deepening mental health crisis, even as regulators acknowledge they haven’t figured out how to ensure these systems are safe for some of the most vulnerable patients. The FDA has authorized 1,524 AI-enabled medical devices as of August 2026. Zero of those clearances cover generative AI products designed to manage mental health conditions. The federal push for AI in mental health Dr. Mehmet Oz, serving as CMS Administrator, publicly endorsed the use of AI avatars for mental health intake in rural areas back in February 2026. His argument was straightforward: staffing shortages in rural mental health care have gotten so bad that an AI chatbot might be better than no therapist at all. The FDA, meanwhile, launched its TEMPO pilot project in late 2025 to evaluate AI tools in real-world clinical settings. The program is specifically designed to collect data on how effective these tools are at treating mental health conditions, with a particular focus on depression. Manufacturers for the pilot are expected to be selected by Septe...