This brief is part of Algorithmikē Psychē, a quarterly series tracking AI policy, research, and mental-health developments. Read the full series on the Algorithmikē Psychē page.
Quarterly digest on AI and mental health — policy, research & society curated by Viveka Mohan Das for Mindful Machines Journal.
Quarter: October 15 – December 31, 2025
Introduction
As 2025 drew to a close, the era of ungoverned AI in mental health care was visibly ending. Three developments — a landmark regulatory hearing, a nationally representative survey, and a wave of state legislation — together marked the quarter as the moment formal accountability arrived. The question was no longer whether artificial intelligence would enter the therapeutic space, but how, and under whose authority.
Quarterly Brief | Q4 2025 (October 15 – December 31, 2025)
This quarter's brief highlights three pivotal moments — each representing a different dimension of an emerging regulatory and public reckoning.
Governance | FDA Convenes Historic Advisory Committee on AI Therapy Chatbots
Source: U.S. Food and Drug Administration / STAT News (November 5–6, 2025)
On 6 November 2025, the FDA's Digital Health Advisory Committee convened its first-ever public meeting dedicated to generative AI in mental health medical devices. The centrepiece was a hypothetical prescription large language model (LLM) therapy chatbot designed for adults with major depressive disorder.
The Committee examined benefits, risks, risk mitigations, premarket evidence requirements, postmarket surveillance, labelling obligations, and integration into clinical care pathways. Experts raised concerns about human susceptibility to AI outputs, the absence of crisis detection safeguards, and the potential for long-term harm from unsupervised use. The FDA Director of Digital Psychiatry at Beth Israel Deaconess Medical Center described the hearing as a signal of the end of "AI exceptionalism" in mental health.
Reference (APA):
STAT News. (2025, November 5). FDA digital advisers confront risks of therapy chatbots, weigh possible regulation. https://www.statnews.com/2025/11/05/fda-digital-advisers-therapy-chatbots-regulating-generative-ai/
Research | One in Four Americans Turns to AI for Mental Health Information
Source: West Health & Gallup (October–December 2025)
A nationally representative survey of more than 5,500 U.S. adults, conducted across the final quarter of 2025, found that approximately 24 percent of respondents had used artificial intelligence tools to explore mental health or emotional concerns — a figure that rose to 13 percent among those aged 18 to 21 specifically seeking mental health support.
The findings, published by West Health in collaboration with Gallup, reflect a rapid normalisation of AI as a first-contact mental health resource. They raise significant clinical and ethical questions about what guidance, if any, these tools are providing, and to whom — particularly given the absence of standardised quality controls.
Reference (APA):
West Health & Gallup. (2025). Millions of Americans now consult AI before, after, and sometimes instead of seeing a doctor. https://westhealth.org/news/millions-of-americans-now-consult-ai-before-after-and-sometimes-instead-of-seeing-a-doctor/
Legislation | US States Begin Codifying Protections Around AI in Mental Health Settings
Source: Psychiatric Times / Manatt Health AI Policy Tracker (2025)
The final quarter of 2025 saw the convergence of multiple state-level legislative efforts to define the limits of AI in mental health care. Illinois enacted protections (effective August 1, 2025) prohibiting AI chatbots from representing themselves as licensed mental health providers. California and Texas passed complementary measures taking effect 1 January 2026: California requiring AI chatbots to disclose their non-human nature to users and to detect and escalate mental health crises; Texas mandating clinician disclosure requirements for AI-assisted care.
Together, these measures mark the first significant legislative architecture for AI in the therapeutic space within the United States.
Read the 2025 psychiatry and AI overview →
Reference (APA):
Miller, B. J. (2025). Psychiatry and artificial intelligence in 2025. Psychiatric Times. https://www.psychiatrictimes.com/view/psychiatry-and-artificial-intelligence-in-2025
Closing Reflection
From regulatory hearing to research survey to statute — Q4 2025 was the quarter in which AI in mental health moved from a speculative frontier to a matter of public policy. The question entering 2026 is not whether guardrails are needed, but whether they will arrive in time to protect the populations already inside the system.
Continue Exploring
Read: Why I Started Mindful Machines Journal
A personal reflection by Viveka Mohan Das on bridging emotional intelligence and artificial cognition.
Explore: Between Brain & Binary — The Explorative Series
A long-form series tracing the intertwined evolution of mental health and artificial intelligence from the 1800s to the present.
Curated by: Viveka Mohan Das
Series: Algorithmikē Psychē | Mindful Machines Journal
© 2025 The Algorithmikē Psychē — Quarterly Brief (Q4 2025)
Continue in this series: Algorithmikē Psychē | Quarterly Brief — Q1 2026. Or return to the Algorithmikē Psychē overview.
Comments
Post a Comment