As director of psychiatric services at PCH Treatment Centers, I have been closely involved in evaluating the use of artificial intelligence in mental health care. Data from leading AI companies and independent research indicate that many people use general-purpose chatbots for help with mental health concerns, and many report finding them useful. These tools are available around the clock and are designed to respond without judgment. Startups are also developing AI therapist chatbots and prescribing agents that are being evaluated in demonstration projects comparable to clinical trials. Other companies have created AI tools for screening and triage. These technologies could expand access to care, streamline the process of arranging treatment, and reduce the administrative burden of clinical documentation, giving clinicians more time and energy for therapeutic interaction. However, excessive chatbot use has also caused harm, including reinforcing delusions, supporting restrictive eating in people with eating disorders, and inadvertently encouraging suicidal or homicidal behavior. AI also raises significant privacy concerns. Drawing on 40 years of psychiatric practice and my role in PCH leadership, I have developed the following Pro-Human Declaration, aligned with PCH’s treatment philosophy:

‍

Artificial Intelligence and Mental Health Care

‍

Psychiatry's foundation is the therapeutic relationship between one human being and another. This declaration is not in opposition to technological progress, but in affirmation of what must never be displaced by it: the irreplaceable humanity at the center of mental health care. AI tools hold genuine potential to reduce administrative burden and broaden access to care. They also risk, if misapplied, deepening isolation, eroding clinical skill, and substituting the appearance of care for its substance.

‍

CORE VALUES:

‍

• Human connection is the therapeutic core. No AI system possesses the consciousness, embodied experience, or relational capacity that grounds authentic psychiatric care. AI must strengthen the clinician's presence and should not replace it.
• AI must not deceive. Patients have the right to know when they are interacting with an artificial system. Transparency about AI identity is a foundational ethical requirement.
• Clinical skill must be protected. Premature AI adoption in training risks producing practitioners dependent on algorithmic outputs without foundational skills to evaluate them. Competence must precede automation.
• AI must not substitute for structural solutions. Technology cannot replace adequate mental health funding, access to trained clinicians, or the restoration of communities fractured by disinvestment.
• Equity and data justice. AI systems reflect existing inequities and may amplify biases related to race, ethnicity, disability, and socioeconomic status. Rigorous, ongoing evaluation across patient populations is required.
• Patient privacy is non-negotiable. Patients have the right to know how their data is used and to opt out of AI-mediated documentation without penalty to their care.