RANZCP Foundation 2026 Psychotherapy Research Grant awarded to Dr Shiang Liew and Dr Steph McLennan

Who Said That? Exploring AI and Psychotherapy

Are psychiatrists safe from the AI boom? In short, opinion is divided and uncertain. Since ELIZA, the first chatbot developed in 1966, artificial intelligence has evolved substantially, with increasingly sophisticated language models now being looked at for use in therapy. For people seeking help, AI may offer a more accessible entry point, with therapy and companionship among the primary reasons people turn to these tools. Yet questions remain about whether AI can be effective, safe and appropriate, particularly when responding to risk, expressing empathy or maintaining relational nuance.

Who Said That? is a pilot study exploring whether people can tell the difference between psychotherapy responses produced by a psychotherapy-trained psychiatrist and those produced by artificial intelligence. Participants will listen to short, recorded therapy scenarios in which both a psychiatrist and AI respond to a clinical vignette. They will then identify which responses they believe came from the psychiatrist, rate their confidence and indicate which response they preferred.

The study uses spoken responses because tone, cadence and timing shape how empathy and authenticity are heard in therapy, yet these elements have rarely been studied. The listening group will include psychiatrists, registrars, mental health clinicians and lived-experience workers, who will listen and respond anonymously.

The study aims to help clarify how AI psychotherapeutic responses are perceived by people, and whether AI may replace clinicians, informing future research, local service planning and safe use in mental health care. So, who said that: “Dr Jones MD FRANZCP” or ChatGPT 5.3?

Read more about RANZCP Foundation-supported projects in the 2025 Impact Report.