Sponsors (3)
Need help making sense of this bill?
This is not an official government website.
Copyright © 2026 PLEJ LC. All rights reserved.
Directs that the Secretary (acting through the Office for Civil Rights) shall receive, investigate, and attempt to resolve complaints of violations of this title in the same manner as complaints of violations of part C of title XI of the Social Security Act (42 U.S.C. 1320d et seq.).
Applies the provisions of section 1128A of the Social Security Act (42 U.S.C. 1320a–7a), except for subsections (a) and (b) and the second sentence of subsection (f), to the imposition of civil monetary penalties under this title in the same manner as those provisions apply to penalties under section 1128A.
Requires health care organizations that use AI clinical decision support systems (AI/CDSS) to keep clinician judgment central: they must adopt written policies, provide training, enable timely clinician override of AI recommendations, and set up oversight and feedback processes with worker representation. The law also bars employer punishment for clinicians who lawfully override AI recommendations, protects workers who report violations, and gives the Department of Labor enforcement authority, civil remedies, and adjustable penalties. Also directs agencies to produce education materials, preserves State law and collective bargaining rights, clarifies malpractice exposure remains for clinicians, limits when override data that identifies clinicians can be shared, and prevents enforcement of predispute arbitration or joint-action waivers for violations of these worker-protection provisions.
Adverse employment action (with respect to a health care professional) includes: (A) termination, suspension, or demotion; (B) any disciplinary action or retaliatory investigation; (C) imposition of a more burdensome work schedule; (D) failure to receive or adverse adjustment in ability to receive a promotion; (E) denial of compensation (including denial of an increase) or any other job-related benefit or opportunity (including telework, training, or travel); (F) revocation of admitting privileges; (G) reassignment of a duty or assignment of an inappropriate duty; (H) change in the ability to practice at a location where the professional otherwise could; (I) an adverse evaluation or performance review; (J) any other modification to terms, conditions, or privileges of employment that, from a reasonable person’s perspective, places the professional in a materially adverse position compared to before the modification; and (K) any other action or inaction that results in a materially adverse position compared to before the action or inaction.
Artificial intelligence clinical decision support system (AI/CDSS) means technology that (A) supports decision-making based on algorithms or models grounded in clinical practice guidelines or that derive relationships from training data (including unsupervised learning models); and (B) produces an output that results in a prediction, classification, recommendation, evaluation, or analysis.
AI/CDSS output means any recommendation, decision, or other output produced by an AI/CDSS.
Commerce; industry or activity affecting commerce — these terms have the meanings given in section 101 of the Family and Medical Leave Act of 1993 (29 U.S.C. 2611).
Covered entity means any individual or entity that (A)(i) employs or otherwise engages a health care professional for remuneration and (ii) is engaged in commerce (including government) or an industry or activity affecting commerce (including government); and (B) includes such an individual or entity that is (i) a health care facility in any setting (for example, a nurse's office in a school) or (ii) a health plan or an administrator of a health plan.
Primary effects
Health care professionals: Gain statutory protections to exercise independent clinical judgment when AI/CDSS provides recommendations, including explicit protection from employer discipline and retaliation when they lawfully override AI outputs. They will likely see new training requirements and opportunities to participate on oversight committees.
Covered entities (hospitals, clinics, health systems, other health care organizations): Must adopt and maintain written policies, provide training, implement feedback mechanisms, form oversight committees with worker representation, and manage data-handling rules for override information. These steps will create administrative and compliance costs and require governance changes.
Patients and public safety: May benefit from stronger clinician oversight of AI recommendations and a formal feedback loop to identify AI performance or safety issues; the law aims to keep clinicians as the final decision-makers for patient care.
Employers and HR/legal teams: Face new compliance and potential enforcement exposure, including civil penalties and private litigation risk if policies or retaliation prohibitions are violated. Predispute arbitration and joint-action waivers cannot be used to block claims under this Act.
Department of Labor and State agencies: DOL gains enforcement duties and will need resources to investigate complaints, issue penalties, and promulgate regulations; States retain enforcement roles and may bring parallel actions consistent with preserved authorities.
AI developers and vendors (indirectly): While the Act focuses on covered entities and worker protections, AI vendors may face increased demands from purchasers for explainability, override logs, and mechanisms that allow clinician override and feedback; purchasers may impose contractual requirements on vendors.
Practical considerations and tradeoffs
Compliance costs vs safety: Organizations will incur costs for policy development, training, committees, and data controls. Proponents would argue these costs are justified by improved patient safety and clinician autonomy; opponents may point to administrative burden and potential liability exposure.
Data privacy and accountability: Protecting clinician identity in override data can encourage appropriate overrides but may limit forensic review in some circumstances; the Act narrows sharing but allows exceptions.
Legal exposure: The Act preserves malpractice liability for clinicians, so overrides do not eliminate civil risk; at the same time, employers face new liabilities if they retaliate or fail to comply.
Labor relations: Worker representation on oversight committees and explicit anti-retaliation protections may strengthen collective bargaining positions and require coordination with unions or worker councils where present.
Read twice and referred to the Committee on Health, Education, Labor, and Pensions.
Introduced October 9, 2025 by Edward John Markey · Last progress October 9, 2025
Expand sections to see detailed analysis
Read twice and referred to the Committee on Health, Education, Labor, and Pensions.
Introduced in Senate