If an AI program is learning from your patient data, are you at risk?

Artificial intelligence (AI) has moved rapidly from the margins of health care into everyday clinical and administrative workflows. Hospitals are using AI systems to flag abnormal imaging findings, predict patient deterioration, automate documentation, manage revenue cycle processes and assist with scheduling and triage. Many physician practices, including smaller independent groups, are beginning to adopt AI-powered transcription tools, clinical decision support systems and patient-facing chat technologies. As these tools become more embedded in care delivery, their influence on medical decision-making, operational efficiency and patient engagement continues to expand.

But with this rapid adoption comes a new and evolving set of legal risks. Questions are emerging about liability when AI systems contribute to clinical decisions, ownership and use of patient data that train algorithms, transparency requirements for AI-driven recommendations, and regulatory oversight that is still catching up to technological change. Health systems and physician practices must also consider risks related to cybersecurity, vendor contracts, algorithmic bias and documentation practices that could affect malpractice exposure or compliance obligations.

Because many AI applications are integrated into existing software platforms, physicians and administrators may not always be fully aware of how these tools operate behind the scenes—or how their use could shape legal responsibility. As adoption accelerates, understanding the legal landscape surrounding AI is becoming as important as understanding the technology itself.

Medical Economics spoke with Dan Silverboard, J.D., a health care attorney with Holland & Knight, to discuss the legal landscape around AI in health care setting.