Roy Perlis: AI, Medical Publishing, and Trust

Generative AI impacts not only how knowledge is generated, but how it is consumed. Ironically, consideration of the real-world impact of big data has been driven largely by anecdote and intuition, with little empirical data to motivate conversation. This presentation will examine, using multiple real-world data sources including population-based surveys, the response to generative AI. It will consider how AI is being used to create and evaluate medical research, and how this may alter both scientists’ and patients’ perception of science.

Roy Perlis, MD MSc is Vice Chair for Research and Ronald I. Dozoretz, MD Endowed Chair in the Department of Psychiatry at Massachusetts General Brigham, where he also directs the hospital-wide Center for Quantitative Health. He is a Professor of Psychiatry at Harvard Medical School and Editor in Chief of JAMA+ AI, responsible for AI content across the JAMA Network. Dr. Perlis has authored more than 500 articles reporting original research in peer-reviewed journals ranging from Nature Genetics and Nature Neuroscience to JAMA and the New England Journal of Medicine, focused on development of novel treatments and biomarkers for brain diseases.

The Perlis lab pioneered the application of emerging artificial intelligence methods to improve patient care in psychiatry, using electronic health records for understanding the course of mood disorders and characterizing treatment response.  Applying these kinds of methods to biobanks allowed his group and collaborators to identify the first risk genes for major depression, in 2016, and for treatment-resistant depression, in 2023. Dr. Perlis was awarded the Depression and Bipolar Support Alliance's Gerald L. Klerman Young Investigator Award in 2010, and the DBSA’s Senior Investigator Award in 2025; he is proud to serve as a scientific advisor to the DBSA. In his spare time, he enjoys tinkering with pinball machines and trying to grow things.