The Future of AI in Mental Health: Opportunities and Challenges

In July 2025, CMS Administrator Mehmet Oz, M.D., MBA, highlighted the potential of artificial intelligence (AI) to revolutionize mental health care accessibility in rural areas. Addressing the shortage of mental health providers, Dr. Oz discussed AI-driven avatars as a promising solution, though the technology remains in its developmental phase. This initiative was reported by Georgia Public Broadcasting.

John Torous, M.D., head of the digital psychiatry division at Beth Israel Deaconess Medical Center, underscored the need for robust studies to validate AI's efficacy in mental health care. He pointed out that leading tech companies like Google and OpenAI have hesitated to claim their AI solutions can provide psychiatric care, focusing instead on wellness support. Dr. Torous noted, "It's telling if none of them claim to deliver psychiatric care with AI."

Previously, Wired reported Dr. Oz's remarks on the cost-efficiency of AI avatars compared to in-person appointments, suggesting AI sessions are significantly cheaper. While patients reportedly rate AI interactions positively, the source of these ratings remains unclear. Mobihealth News further detailed Dr. Oz's suggestion of AI's role in initial mental health assessments, particularly its ability to detect subtle communication cues.

Regulatory and Legislative Implications

Despite AI's potential in supporting mental health services, substantive evidence-based research is still required. An initiative in the UK demonstrated benefits where people with schizophrenia used avatars to alleviate distress. However, AI's limitations in healthcare triage and therapeutic discussions persist. Dartmouth's study in NEJM AI showed AI chatbots effectively alleviating certain conditions, yet the study lacked a direct control for comparison.

Concerns about AI in mental health include associations with adverse outcomes like suicides linked to chatbots, prompting legal actions against companies such as OpenAI. Carrie Henning-Smith, Ph.D., from the University of Minnesota, warned of the impacts of unverified tech use on vulnerable populations.

Regulatory changes are underway, with Illinois enacting legislation limiting AI use in behavioral health without expert certification among over 250 related proposals across states. Colorado introduced laws this year focusing on child safety and suicide prevention. The insurance industry should monitor these legislative trends to assess impacts on AI healthcare integration.

Journalists are encouraged to explore research and state-level adaptations under CMS’s rural health initiatives. A notable example is East Carolina University's comprehensive program utilizing telepsychiatry for youth mental health screenings in rural North Carolina, set to be presented at the 2026 HIMSS Global Health Conference & Exhibition.