Since ChatGPT was released in December of 2023, there has been a lot of conversation about how artificial intelligence may upend various facets of human life. From academia to the arts, from the media to the medical field, the effects are far-reaching and serious, whether you take the perspective that they are exciting, terrifying, or both.
Meanwhile, the field of psychotherapy has often been assumed to be one of the latter ones that will be affected. After all, isn’t it the foundation of human connection? Yet, it turns out that this isn’t the case.
Evidence shows that young people, in particular, are turning not just to therapy bots but to more general AI, such as ChatGPT, for help. One study (McBain et al, 2025) found that 1 in 8 adolescents had used AI chatbots specifically for mental health advice. This has raised some serious concerns, including AI’s potential support of suicide for vulnerable individuals, and the way it can accelerate psychosis through its sycophantic and flattering responses to those who are growing out of touch with reality.
What has received less attention, however, is how AI is creeping into person-to-person therapies — sometimes unbeknownst even to the therapist.
Therapists and AI
Automated note-taking software, transcription services, recording platforms, and scheduling services for therapists all promise improved efficiency and streamlined administrative tasks. And while those sound wholly positive, the agreements that therapists enter into may not clearly elucidate potential privacy risks and the long-term plans of these companies. What do these platforms do with clients’ data? How do they protect confidentiality? Who owns the data? Is the data ultimately being used to train the therapists’ robot replacements?
Therapists themselves may also use AI for consultation (help with difficult cases), just as in any other field where expertise is sought. But this raises confidentiality concerns as well. Even conscientious therapists might not be well-versed in the risks. And how many of them raise uneasy questions ranging from (at best) how easily a client’s privacy can be compromised to (at worst) whether a client’s data is being used to train therapy bots that will eventually replace the therapists themselves?
7 questions to ask your therapist about their use of AI
With so many uncertainties, seeking clarity is paramount. As a consumer, you have the power to initiate this conversation and get answers about exactly what is happening to your personal story.
The first step is to thoroughly read the informed consent document that you signed when you started therapy. If you no longer have a copy, ask for a new one. Then, consider asking your therapists these questions:
- Has anything changed in the way you manage your practice since I signed the informed consent and practice policy documents?
- In what ways do you use artificial intelligence in your practice?
- Do you use AI for notetaking, transcription, or case summaries?
- If we do telehealth, are my sessions recorded? Are they monitored in real-time by any platform?
- Do you use AI for treatment decisions?
- How is my privacy ensured in each of these steps?
- Will I be notified if you do start using software that is AI-related?
It can be intimidating to seek therapy, and clients may often not want to rock the boat by asking questions about something that may feel unrelated to their treatment. But these questions are actually central to treatment. And all of us — therapists, clients, and the general population alike — should be clear about how our lives’ most private details are being seen and used.
