From ClingantryAI Health Regulator News

Georgia's SB 444, effective January 1, 2027, draws a clean line: insurers may use AI in prior authorization to automate paperwork and reduce administrative burden — but an adverse decision (denying or reducing care) can't be issued until a qualified clinical peer has actually conducted the review and participated in the decision. The AI also can't override that clinician's judgment. If you operate in Georgia, nothing changes today, but this is worth having on your 2027 compliance calendar now.

Utah's SB 319, in effect since May 6, 2026, requires insurers to publicly post their preauthorization requirements on their website and to disclose whether AI is used when reviewing authorization requests. It also sets a maximum time insurers can take to answer an authorization request, and sets minimum periods an approval stays valid. Practical upshot for your utilization and scheduling teams: approvals should be faster to get and harder to have silently expire — and you can now find out whether a Utah payer is using AI on your requests just by asking.

As of July 1, 2026, health insurers in Indiana can no longer use AI as the only basis to downcode a claim (pay it at a lower level than billed) — a human must review the patient's medical record first. Insurers also now have to disclose, in plain view, whenever AI was used to deny a prior authorization or downcode a claim. Two things for your team: your billing office should start watching downcoded claims for the required disclosures, and note the law cuts both ways — providers also can't submit claims generated by AI without a person involved in the claim reviewing them. Appeals get at least 180 days.

Alabama's new law (signed April 17, 2026, effective October 1, 2026) says a health insurer can't rely only on AI to decide whether care gets covered — any decision to deny or reduce coverage has to be made by a qualified healthcare professional. Insurers must also tell enrollees that AI is used in coverage decisions, base prior-auth determinations on the individual patient's medical history rather than group data alone, and certify annually to the state that their AI is monitored for accuracy and doesn't discriminate. If you're in Alabama, this is leverage: after October 1, an AI-only denial is something you can push back on by law.