Elizabeth Matthews

Bio:
Dr. Matthews is an Associate Professor at Fordham University's Graduate School of Social Service. Her research broadly focuses on communication and decision-making in health and mental health, with a particular emphasis on how technology intersects with clinical practice. She has recently focused on the implementation of AI in mental health services, specifically large language models and AI clinical documentation. Her work has been supported by the Robert Wood Johnson Foundation, the Health Resources and Services Administration, and the National Association of Social Workers. 

Abstract:
AI’s technical capabilities in mental health are rapidly proliferating. Yet the central question for trustworthy AI is not only what AI can do, but what it should do. In mental health, this requires addressing a fundamental challenge: defining “ground truth.” Unlike domains with objective metrics, mental health has long struggled to establish consistent, discrete parameters for diagnosis, assessment, and treatment. The historical ambiguity is now colliding with AI's need for objective data, and risks building tools on an unstable foundation. Resolving these challenges requires leadership, yet a lack of consensus from practitioners and regulators on AI's role in mental health is creating an environment of uneven, unsupervised implementation The path forward is not simply about what AI can do, but deciding, together, what AI should do: build on trusted standards, reflect the voices of clinicians and patients, and serve as a tool to enhance human judgment.