Sponsors (7)
The bill strengthens patient protections, transparency, clinician oversight, and parity for mental-health/substance-use coverage where AI is used — but at the cost of higher administrative burdens, slower decisions, privacy/proprietary risks, and potential reductions in AI-driven efficiency that could raise costs or narrow benefits.
Patients — especially people with chronic conditions and beneficiaries on Medicare and Medicaid — will have adverse clinical-judgment coverage denials reviewed and signed by a licensed clinician rather than solely by AI, preserving human oversight and reducing risk of inappropriate denials.
Enrollees will get clearer transparency: when an AI system influenced a coverage decision plans must notify the enrollee, describe the AI's role, and identify the clinician (name and credentials), improving the ability to appeal and understand decisions.
People with mental health and substance-use needs (and related beneficiaries including many on Medicaid) gain stronger parity protections because AI-driven utilization review is treated as a treatment limitation, helping prevent covert or automated reductions of MH/SUD benefits.
Health plans, vendors, and ultimately consumers and employers will face higher administrative and compliance costs to ensure clinician review, document AI systems/outputs, and produce disclosures — costs that could raise premiums or shift to taxpayers, employers, and enrollees.
Patients — particularly those with chronic conditions — may experience slower coverage determinations and longer internal appeals because clinicians must perform independent evaluations rather than deferring to automated decisions.
Insurers and third-party reviewers may reduce or limit use of automated AI tools to comply with the rules, eroding efficiency gains and innovation in utilization review and care management and disproportionately affecting smaller plans that rely on automation.
Based on analysis of 5 sections of legislative text.
Requires licensed clinicians, not AI systems, to make adverse coverage decisions involving clinical judgment and mandates notice, parity disclosure, and recordkeeping when AI is used.
Official title: To amend title XXVII of the Public Health Service Act, the Employee Retirement Income Security Act of 1974, and the Internal Revenue Code of 1986 to regulate the use of artificial intelligence in the review of claims by group health plans and health insurance issuers offering group or individual health insurance coverage.
Introduced September 1, 2026 by Greg Landsman · Last progress September 1, 2026
Prohibits artificial intelligence systems from issuing or dictating adverse health insurance benefit decisions that involve clinical judgment and requires those decisions to be made by licensed health care professionals who provide independent clinical judgment. Requires plans to disclose when AI materially influenced plan design, utilization review, or quantitative treatment limitations, to provide specific notices to enrollees when AI was used, and to keep administrative records describing the AI system, its role, outputs, and compliance. Applies the same requirements across the Public Health Service Act, ERISA, and the Internal Revenue Code for group and individual health coverage. Implements a delayed effective date so the rules apply to plan years beginning on or after the January 1 that starts at least 12 months after enactment.