
Agentic AI in Healthcare: HIPAA Liability, Security Risks & Regulation
What happens when AI stops answering questions and starts taking action? Who is responsible when an autonomous AI agent makes a mistake involving patient data? In this episode of the HIPAA Insider Show, host Adam Zeineddine welcomes back Sam Hart, CEO of Hathr.ai, to discuss the rise of agentic AI in healthcare, the risks of giving AI systems write access, and the growing questions surrounding HIPAA liability and AI regulation. Unlike traditional Large Language Models (LLMs), which primarily generate responses, agentic AI can execute workflows, interact with applications, and potentially modify sensitive healthcare records. As AI becomes more autonomous, healthcare organizations must rethink security, accountability, and patient data protection. IN THIS EPISODE: Agentic AI vs. Generative AI: What changes when AI moves from answering questions to executing tasks? AI Write Access: The risks of allowing autonomous agents to modify healthcare systems and patient records HIPAA Liability: Who is responsible when an AI agent mishandles Protected Health Information (PHI)? AI Regulation: The debate over FDA oversight and the roles of HHS and the Office of Science and Technology Policy (OSTP) Autonomous AI Security: How machine-scale actions can amplify security and compliance risks AI Governance: Why access controls, human oversight, and compliance vetting matter The big question: If AI agents can independently access and modify healthcare data, how do organizations maintain control and accountability? Whether you're a healthcare IT professional, digital health founder, compliance officer, or AI developer, this conversation explores the opportunities and challenges of deploying autonomous AI in healthcare. LEARN MORE ABOUT HIPAA VAULT
