Sponsors (3)
The bill strengthens clinicians' authority, transparency, and enforcement around clinical AI—improving patient safety and worker protections—but does so by imposing substantial compliance, litigation, and financial burdens on providers, developers, and state actors while leaving malpractice exposure intact for clinicians.
Health care professionals (employees and non-employee clinicians with admitting privileges or similar arrangements) are protected from materially adverse employment actions for overriding AI/CDSS outputs and retain final clinical authority to override AI recommendations, preserving clinical judgment and patient safety.
Patients and clinicians get clearer transparency about when AI is used and how—providers must disclose AI use, inputs, and limitations—helping patients understand care and clinicians explain decisions.
The bill defines key terms (AI clinical decision support, AI/CDSS output, override, override data) giving providers and developers clearer compliance boundaries when deploying AI in care settings.
Hospitals, clinics, health plans, and other covered entities will face substantial new administrative and compliance costs (tracking override data, creating policies/committees, training, reporting, recordkeeping), increasing operational burdens and potentially shifting resources away from care.
Clinicians, developers, and employers face increased litigation and liability risk because granular logging and possible disclosure of override data (including in malpractice, negligence, or law-enforcement actions), exposure to statutory/treble damages, and greater incentives to sue could raise insurance and legal costs.
The bill creates large civil penalties and statutory damages for violations (per-violation fines and repeat penalties), which could impose significant financial liabilities on providers and insurers.
Based on analysis of 9 sections of legislative text.
Requires policies, training, oversight, and privacy limits so clinicians can override AI clinical decision systems without retaliation; creates HHS and DOL enforcement.
Official title: Protect the independent judgment of health care professionals acting in the scope of their practice in overriding AI/CDSS outputs, and for other purposes.
Introduced October 9, 2025 by Edward John Markey · Last progress October 9, 2025
Requires health care employers and other covered entities that use artificial intelligence clinical decision support systems (AI/CDSS) to adopt policies that preserve clinician independent judgment, permit timely clinician overrides for patient-appropriate or legally necessary reasons, and limit sharing of clinician-specific override data. It creates training, notice, and committee-review requirements, protects clinicians from retaliation for overriding AI outputs or reporting violations, and assigns enforcement to HHS (civil penalties) and the Department of Labor (labor/whistleblower enforcement), while allowing certain data disclosures to patients and for legal proceedings.