Focus Groups Reveal Deep Caution Among Experts on AI Chatbot Therapists

AI chatbots are increasingly being used as therapists for people with mental health problems, and a new qualitative study from the Netherlands finds that the professionals best placed to judge the technology remain deeply wary. Researchers convened online 90-minute focus groups with mental healthcare practitioners, including psychiatrists, psychologists, psychotherapists, physicians and psychiatric nurses, alongside ethicists, AI experts and legal professionals, to examine the ethical questions raised by chatbots acting in a therapeutic role. Across every stakeholder group, participants expressed caution or outright reluctance toward using AI chatbots in mental healthcare, citing unresolved ethical and legal concerns.

The study, published in BMC Medical Ethics, was guided by a topic list based on a previously published scoping review, and a vignette served as a prompt for ethical reflection. The discussions were analysed using thematic content analysis. Notably, the researchers refined their methods as the study progressed to enhance validity and reliability, adjusting the topic list, the vignette or the group composition along the way. Transcripts were reviewed, corrected and coded in MAXQDA, a qualitative data analysis programme.

The worries participants raised were concrete rather than abstract. Key concerns included potential harm to users, uncertainty about who would be responsible for adverse outcomes, and risks such as harmful output or AI hallucinations, excessive screen time and limited effectiveness.

Human contact emerged as another dividing line. While most participants preferred human-to-human contact in therapy, they acknowledged that some patients might prefer chatbots, and they emphasised the importance of informed and autonomous choice in any use of the technology. Privacy and confidentiality concerns were also raised, but interestingly these came particularly from legal professionals rather than from the other stakeholder groups, suggesting that different professional lenses surface different ethical risks.

One of the study’s most useful contributions is a taxonomy of ethical concerns. The researchers distinguished three types: general AI issues that apply wherever the technology is deployed, concerns specific to AI in mental health contexts, and concerns that also apply to human therapists. Distinguishing between these categories matters, the authors argue, because potential solutions may come from different domains.

Participants weighed the second category, concerns specific to AI in mental health contexts, especially heavily. Yet the authors also point out that people are already turning to chatbots for mental health support, regardless of professional reservations. That reality reframes the debate. Rather than asking whether AI chatbots should be used in mental healthcare, they suggest the real question is under what conditions, and with what level of risk acceptance, their use can be responsible and safe.

The study’s motivation rests partly on a practical asymmetry: unlike human practitioners, chatbots are not affected by waiting times, which are a persistent bottleneck in mental healthcare systems. The study itself does not test whether chatbots are effective or safe; it documents how a diverse group of stakeholders reasons about the ethics of their use, which is a necessary precursor to any responsible deployment.

As qualitative research with participants in the Netherlands, the findings reflect the perspectives of specific professional groups in a particular healthcare system, and the authors note that methods were deliberately refined during the study to strengthen validity and reliability. Even so, the convergence of caution across ethicists, clinicians, AI specialists and lawyers is striking.

Subject of Research: Ethical concerns about AI chatbots used as therapists in mental healthcare, examined through multi-stakeholder focus groups in the Netherlands

Article Title: Ethics of AI chatbots in mental healthcare: insights from a multi-stakeholder focus group study

Article References: Rahsepar Meadi, M., Jaspers, F., Metselaar, S., Batelaan, N., & Bernstein, J. (2026). Ethics of AI chatbots in mental healthcare: insights from a multi-stakeholder focus group study. BMC Medical Ethics. https://doi.org/10.1186/s12910-026-01643-9

Image Credits: AI Generated

DOI: 10.1186/s12910-026-01643-9

Keywords: AI chatbots, mental healthcare, medical ethics, focus groups, qualitative research, therapists, responsibility, privacy, informed consent, Netherlands, BMC Medical Ethics, stakeholders