University of Pittsburgh AI Healthcare Research Symposium Poster

AI Accountability in FDA Regulated Clinical Trials

Artificial intelligence (AI) is rapidly transforming FDA-regulated clinical trials, offering advancements in patient selection, adverse event monitoring, and predictive analytics. However, the integration of AI introduces significant ethical, legal, and regulatory challenges, particularly concerning accountability when errors occur. This presentation examines key concerns such as algorithmic bias, lack of transparency in AI decision-making, and the absence of a defined liability framework within the current FDA regulatory structure. Using case studies—this research highlights the risks of unvalidated or biased AI models. A comparative analysis of U.S. and European regulations reveals contrasting approaches to AI oversight. The presentation concludes with proposed policy recommendations aimed at enhancing transparency, mitigating bias, establishing shared liability, and reinforcing human oversight in AI-supported clinical research.