Trustworthy Conversational AI Agents for Mental Health Support
Date:
In this talk, we present a concrete research example from the domain of mental health support. Recent studies show that patients often appreciate conversational AI for its accessibility, affordability, and non-judgmental character. At the same time, using AI in such sensitive contexts raises critical challenges around trustworthiness, reliability, and ethical responsibility. Drawing on several recent and ongoing projects, we explore how conversational agents can be designed to prioritize these concerns, and we reflect on the importance of collaboration between psychotherapy practitioners, researchers, and HCI/AI experts to advance responsible innovation.