Iowa recently introduced legislation mandating that patients must give consent before artificial intelligence (AI) can be utilized to record and transcribe clinical interactions. In a separate development, Utah has passed a law asserting that AI does not fall under the category of innovation or technology enhancements within the medical practice environment.
The following outlines the newly implemented measures for 2026, organized by relevant areas of concern.
**Restricting AI in Medical Authorizations**
These recent legislative measures respond to growing apprehension about the deployment of AI systems by healthcare insurance providers, which can automatically grant or deny approval for medical treatments.
Most of the legislation in this area requires that a human medical professional be involved in any decision-making regarding authorizations. Additionally, decisions must reflect the nuances of the patient’s medical history and individual circumstances, rather than relying exclusively on aggregated data.
**Alabama**
SB 63 pertains to the role of AI in healthcare insurance authorizations. This law establishes regulations around the use of AI by insurers for prior authorization decisions. It mandates that insurers must be transparent about their use of AI and that any AI-generated decisions must be informed by the specific medical history and clinical conditions of the patient. Furthermore, any denial of coverage must involve a licensed physician or healthcare professional. This legislation, sponsored by Orr, was enacted on April 17, 2026, and is set to take effect on October 1, 2026.
**Colorado**
The bill HB 1139 addresses the regulations surrounding AI use in healthcare authorization choices. It requires insurers to ensure that medical authorization decisions account for a patient’s unique medical history and clinical circumstances, avoiding reliance on general data alone. In the event of a coverage denial, a licensed clinician or physician is obligated to review the case. Sponsored by Joseph and Lieder, this legislation was enacted on June 2, 2026, and will be effective starting January 1, 2027.
**Georgia**
Georgia’s SB 444 establishes a prohibition against making insurance coverage decisions based exclusively on AI systems or software. The bill was sponsored by Kirkpatrick and others, and it was enacted on May 5, 2026, with an effective date set for January 1, 2027.
**Illinois**
The legislation HB 3114 in Illinois restricts healthcare payers from adopting any policy or applying any algorithm or automated process that fails to consider the complete information provided by billing healthcare professionals when downcoding claims. The bill allows the use of automated systems to flag potential downcoding opportunities; however, any final determinations must be carried out or reviewed by a human. This bill, sponsored by Koehler and others, is currently awaiting the governor’s approval.
**Iowa**
Iowa's HF 2635 focuses on standards of conduct for health insurance. It permits insurers to leverage AI for the initial review of prior authorization requests, but decisions to deny, delay, or downgrade such requests cannot solely rely on AI. This law was enacted on May 13, 2026.
**Utah**
Utah's SB 319 updates preauthorization requirements for health insurance to include stipulations on AI usage. Insurers are required to publish preauthorization details on their websites and disclose AI's role in authorization reviews. This bill mandates that individuals evaluating adverse preauthorization decisions apply independent medical judgment, rather than depending solely on external recommendations. Additionally, it sets forth time limits for authorization determinations, outlines minimum durations for chronic or long-term care authorizations, and requires insurers to provide information on adverse rulings. Sponsored by Johnson and Hall, the law was enacted on March 19, 2026, and will take effect on January 1, 2027.
**Washington**
Finally, Washington’s SB 5395 enhances existing regulations to ensure transparency and accountability concerning the use of AI in prior authorization processes. The new law clarifies that only licensed physicians or health professionals may deny prior authorization requests based on medical necessity. AI is not permitted to serve as the sole basis for denying, delaying, or altering healthcare services. Sponsored by Orwall, Muzzall, and others, this law was enacted on March 23, 2026, and is effective as of June 11, 2026.
