Healthcare AI Guardrails

State-Level Oversight

• California enacted three healthcare AI measures. AB 1979 limits AI to an advisory role in clinical functions, SB 503 requires efforts to identify and mitigate foreseeable bias, and AB 1609 restricts some chatbot claims to be human; the package also expands protections for health information accessed through chatbots and requires documentation of intended uses, risks and bias controls.

Healthcare IT NewsMedCity News

• Pennsylvania’s House passed a healthcare AI disclosure bill, 144–59. HB 1925 would require facilities to disclose AI use in clinical decisions, explain it in plain language online, identify AI-generated patient communications and preserve human judgment—giving patients clearer notice while leaving the proposal short of a statewide standard.

Tri-State Alert