Shalaila spoke at the PsychE / PsychAI Symposium 2026

Created on April 19, 2026

2026

Dr. Shalaila S. Haas gave an invited talk, “AI and Psychotherapy: Insights from the PREDiCTOR Study,” at the PsychE / PsychAI Symposium 2026 on April 18, 2026. Held at Mount Sinai’s Goldwurm Auditorium, the hybrid event brought together a multidisciplinary group of clinicians and researchers to discuss the intersection of artificial intelligence and mental health care.

In her presentation, Dr. Haas addressed the urgent need for precision medicine in psychiatry to navigate the high rates of treatment dropout and resistance. She highlighted how the PREDiCTOR Study (Phenotypes REimagined to Define Clinical Treatment and Outcome Research) is working to equip clinicians with “behavioral vital signs”—objective markers derived from AI analysis of clinical encounters.

Key Highlights from the Talk:

  • Objective Measurement: Dr. Haas explained how AI can quantify aspects of clinical conversation that are nearly impossible to track manually, such as speech acoustics (vocal pitch and prosody), facial expressions (frame-by-frame muscle movements), and natural language patterns (coherence and emotional tone).
  • The PREDiCTOR Framework: The study utilizes six data streams—including audio/video recordings, smartphone passive data, and electronic health records—to follow 2,100 patients over one year.
  • Predicting and Preventing Crisis: A major goal of the research is to create an early warning system for clinicians to prevent treatment disengagement, ER visits, and hospitalizations before they occur.
  • Patient-Centered Technology: Dr. Haas shared results from testing various recording setups, noting that patients felt most comfortable and least distracted by the OBSBOT Meet 2 webcam compared to more obtrusive smartphone setups.
  • The Power of Therapeutic Alliance: Preliminary findings showed that therapeutic alliance (TA) scores significantly impact outcomes; patients who later experienced a crisis event or dropped out often showed lower TA scores in initial sessions.

Dr. Haas concluded by emphasizing that AI is not intended to replace the clinician, but rather to provide them with tools to “listen more deeply” and hear what the human eye and ear might miss in the pursuit of personalized care.