Electronic health records (EHRs) are a central data source for artificial intelligence (AI) in healthcare as they contain rich data on large populations of patients and the medical care they receive. AI with EHR data has already accelerated biomedical research to improve patient care, including studies uncovering the effectiveness of medical treatments and new tools that predict disease risk and progression. However, the analysis of EHR presents numerous challenges that have bottlenecked the widespread and responsible use of AI in medical practice. The objective of this workshop is to bring together international leaders across academia, industry, and government in a collaborative atmosphere to tackle the most pressing challenges in responsible AI with EHR data.