As healthcare AI enters 2026, the industry stands at a critical crossroads. One path emphasizes rapid deployment of generative AI tools, such as ChatGPT Health, promising scale and efficiency. The other urges caution, as academic leaders at Stanford and Harvard warn of safety risks, hallucinations, bias, and unresolved ethical concerns. Based on systematic evaluation of leading LLMs, current evidence suggests GenAI is not yet ready for routine clinical care without rigorous validation and continuous monitoring. Without new approaches to accelerate clinical evidence generation at scale, healthcare AI risks repeating past boom-and-bust cycles rather than achieving sustainable, equitable impact.


At the Chen Institute’s AI Accelerated Science Symposium on October 28, 2025, ELHS Institute founder Dr. AJ Chen proposed a new vision for Open Clinical AI Science (OCAIS) to accelerate the clinical impact of generative AI. The framework delivers free GenAI-based disease prediction services to clinical teams worldwide, including low-resource settings, enabling large-scale participation in clinical evidence generation. By converging GenAI with task-specific Learning Health System units, this approach aims to shorten evidence-generation timelines from decades to years and help prevent GenAI from repeating past failures in health care innovation.
Dr. AJ Chen delivered a keynote at the Tsinghua Health AI Summit on converging generative AI (GenAI) and Learning Health Systems (LHS) to improve clinical diagnosis and reduce global health disparities. He presented the ELHS Institute’s ML-enabled LHS framework, supported by Nature- and JAMIA-published studies, showing how GenAI embedded in LHS units can enable scalable, responsible evidence generation and democratize high-standard predictive care worldwide.
