At the 2025 Aegis Summit, Eric Schadt, PhD, Chief Scientific Officer at Pathos and Professor/Dean for Precision Medicine at Icahn School of Medicine at Mount Sinai, delivered a deep and provocative look at how AI can fundamentally change drug discovery—not by generating more molecules, but by transforming how we decide which patients should receive which drugs.
Eric argued that despite extraordinary advances in AI, genomics, and imaging, drug labels in 2025 still rely on overly simplistic, single-variable biomarkers. He made the case for moving beyond “single-needle” predictors toward multi-modal, fusion-sensor models that integrate the rich clinical data already sitting in medical records. Drawing on large-scale clinical trial and real-world validation, he showed how AI-driven, mechanistically informed biomarkers can dramatically outperform today’s standards and improve patient outcomes.
The talk closed with a call to action: realizing AI’s promise in drug development will require trusted collaboration across health systems, payers, regulators, and investors—building the evidence, incentives, and pathways needed to bring AI-earned labels into clinical care.
