Roman Malo, Ph.D.

Roman’s interest in AI is neither a fad nor a technophile pose. It is part of a long inquiry into what emerges from an encounter between a subject and a device: video games in adolescence, virtual reality during his thesis, conversational AI today. One and the same question structures this triptych: how does something like a consciousness emerge in contact with a device?

“What interested me in psychology was trying to understand how consciousness emerges, the problem of consciousness, which for me is always a fascinating subject. […] I did little trial-and-error experiments on ChatGPT and Claude, like many people. I tried to push the technology quite far: I went from the cooking recipe to really pushing a line of reasoning, to see what happened. I found it fascinating.”

This experimental stance leads him to reject from the outset two figures he frequently encounters among his colleagues: technophile fascination that projects the divine onto the tool, and defensive rejection that reduces it to “an improved calculator.” For Roman, conversational AI is radically new: it raises fundamental anthropological questions about agency, autonomy, consciousness and solitude that neither classic computer science nor academic psychology are prepared to address alone.

A long testimony collected in May 2026 for this site develops these analyses as well as his reading of AI in university teaching, his hypothesis of the polarization of consultations, and his use of AI as a pedagogical prosthesis. This document is currently being reviewed by Roman before publication.

Roman Malo holds a singular position in the French-speaking landscape of researchers working at the intersection of clinical work and technology. His profile matters to practitioners for four reasons.

• Critical reading of AI research: his competence in data ecology provides concrete tools to evaluate whether a study on a therapeutic chatbot says something about your practice — or only about laboratory conditions.
• A bridge between technology and clinical work: trained at once in clinical psychology, human-machine interaction and phenomenology, he thinks about technological devices from the patient’s experience — not from engineer’s metrics.
• A transdiagnostic approach: his process models (psychological flexibility, network models) apply across the board — not limited to a specific disorder. This is valuable for a non-reductive evaluation of AI in psychotherapy.
• Micro-phenomenology: his use of the explicitation interview to capture lived experience in virtual reality offers a methodological model for studying patients’ experience with chatbots — beyond simple satisfaction scoring.