0
0
A recent confrontation between a mental health patient and their therapist in Los Angeles has exposed a growing crisis of trust in the mental health sector, as therapists increasingly—and quietly—turn to AI chatbots like ChatGPT during therapy sessions. The incident, part of an emerging trend documented in investigative reports and supported by expert commentary, sparks ethical concerns around transparency, privacy, and the integrity of therapeutic relationships as artificial intelligence finds new roles in mental health care.
The Unfolding Trust Crisis
Stories are emerging of patients discovering that their therapists have used AI, such as ChatGPT, to guide conversations, compose emails, and even generate prompts during therapy sessions—often without disclosure. One patient, Declan, described his therapist’s admission of using ChatGPT as a “weird breakup,” made worse when the session was both emotionally charged and still billed. Another, Hope, felt uneasy after recognizing an AI-generated message in a consolatory email, underscoring the trust issues that brought her to therapy in the first place.
Investigative reporting by MIT Technology Review and other outlets has revealed that such practices are not isolated, but symptomatic of a potential shift in how mental health care is delivered. The foundational elements of psychotherapy—authenticity, empathy, and a safe space for vulnerability—are threatened when patients feel “copy-pasted” by their provider.
Key Findings and Developments
Recent surveys of community members (CMs) and mental health professionals (MHPs) suggest that around 28% of patients and 43% of professionals are using AI tools in some capacity. While benefits are reported in accessibility and administrative burdens, nearly half of users—both patients and professionals—have experienced harms such as reduced human connection and concerns about accuracy, data security, and ethics.
Therapists report using AI for session guidance and note-taking, but the lack of transparency and proper safeguards remains a problem. Some platforms are now offering HIPAA-compliant AI tools, but most general-purpose chatbots, including ChatGPT, do not meet these privacy standards, putting confidential patient data at risk.
Expert Perspectives
Adrian Aguilera, PhD, clinical psychologist at the University of California, Berkeley, warns, “People value authenticity, particularly in psychotherapy. Using AI without disclosure can feel like you’re not taking the relationship seriously.” Dr. Pardis Emami-Naeini, Duke University, highlights privacy risks: “General-purpose AI tools like ChatGPT are not HIPAA compliant, which means patient information could be at risk.”
Vaile Wright, PhD—from the American Psychological Association—noted recently, “AI tools can help expand access to care but must never compromise the core values of personal connection, confidentiality, and informed consent.”
Context: Why AI Is Entering Therapy
Burnout among mental health professionals, rising demand for therapy, and shortages in access have made AI appealing as a clinical support tool. AI can automate note-taking, synthesize responses, and handle administrative tasks. It can help therapists manage workloads more efficiently, freeing up time for direct patient care.
Platforms such as Heidi Health and Upheal now market HIPAA-compliant AI modules for therapists, emphasizing encrypted data storage and secure communications. However, these are relatively new and adoption remains limited compared to general chatbots.
Implications for Public Health
The implications are profound. Therapy is a trust-dependent process, and breaches can stall or reverse progress in a patient’s mental health journey. Uninformed use of ChatGPT or similar tools raises risks of misdiagnosis, inappropriate advice, and data exposure. About 47% of surveyed community members and 51% of professionals reported specific harms or concerns tied to AI use, emphasizing accuracy, personalization, and ethics.
Concerns extend to crisis scenarios: AI chatbots are not yet reliable in recognizing imminent risk, such as suicidal ideation, and may inadvertently escalate harm if misused. Regulatory bodies advise that any use of AI in mental health should come with clear disclosures and stringent safeguards.
Potential Limitations and Counterarguments
AI in therapy is not intrinsically harmful. Some studies show that clinicians cannot always distinguish between well-written AI-generated advice and human responses. AI can offer benefits: expanded access, lower costs, and support for marginalized populations. Nevertheless, the technology is in its infancy for direct clinical use, and notable failures—including inappropriate advice to vulnerable patients—underscore the need for cautious integration, better oversight, and training for providers.
Experts argue that transparency—informing patients when AI tools are used—and rigorous compliance with privacy laws are fundamental for ethical care.
Practical Implications for Readers
-
For patients: If therapy feels scripted, ask your therapist about the role of technology in your sessions. Informed consent is your right.
-
For professionals: Transparency, documentation, and ongoing training are essential. Select only privacy-compliant AI tools and regularly review best practices.
-
For the public: The rise of AI in mental health care signals promising opportunities but also significant challenges. Watch for new guidelines from trusted organizations.
Medical Disclaimer
Medical Disclaimer: This article is for informational purposes only and should not be considered medical advice. Always consult with qualified healthcare professionals before making any health-related decisions or changes to your treatment plan. The information presented here is based on current research and expert opinions, which may evolve as new evidence emerges.
References
- https://health.economictimes.indiatimes.com/news/industry/trust-crisis-in-mental-health-patient-exposes-therapists-use-of-chatgpt/123736365?utm_source=top_story&utm_medium=homepage
