Original title: Understanding Public Trust in AI-Based Clinical Psychological Decision-Making: A Qualitative Study
Yangqiong Lei, Samuel Bennett · September 2026
Artificial intelligence can help healthcare workers make decisions and work faster. However, people worry about using computer programs for mental healthcare, where private details are sensitive and personal. To see what shapes public trust, researchers interviewed people using open-ended questions and looked for key patterns in their answers. They found that trust relies on several major elements: how clear the tool is, how it is designed, fairness, privacy, rules, and who takes the blame for errors. People also care about how the system works with humans. They trust AI more when it assists a human doctor rather than making choices alone without a person involved. Finally, public knowledge about healthcare and technology shapes acceptance. For people to trust these systems, developers must protect privacy, prevent unfair biasBias means a system may treat certain groups unfairly because of flaws in its training data or design., establish clear laws, and help the public learn how these medical tools work alongside human professionals.
You're planning a family trip and a travel app suggests an itinerary. If you can tweak it and the app explains its reasoning, you're happy to use it. But if the app books everything on its own and you can't see why, you'd worry it missed something important. In this paper, people felt the same about a mental-health tool: they trusted it more when it helped a professional decide than when it decided on its own.
AI is moving into mental health care, where decisions carry real weight. If the public doesn't trust AI-based diagnoses or treatment recommendations, people may reject tools that could make care faster and more accessible. This study points to what builds that trust — transparency, privacy, regulation, and keeping a human in the loop.
The abstract doesn't state limitations explicitly. As a qualitative interview study, it does not establish cause and effect or measure how widespread these views are. The sample is not described, so the findings may not represent the broader public.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Understanding Public Trust in AI-Based Clinical Psychological Decision-Making: A Qualitative Study — Yangqiong Lei, Samuel Bennett — PsyArXiv — doi:10.31234/osf.io/u96y5_v1
Abstract licence: CC BY
Explanation generated by Synthesis from the abstract — it is not the original paper. This is a preprint, so it hasn't necessarily been peer reviewed.