When AI Denies Your Claim, Alabama Says a Doctor Has to Own It

Alabama's Senate just passed a bill that does something simple but significant: if a payer uses AI in prior authorization and the decision is a denial, a physician has to sign off. Not an algorithm. A doctor. If it passes the House, it takes effect October 1.

The bill passed the Alabama Senate on February 19 and has been referred to the House Insurance Committee. It would apply to any entity issuing, delivering, or renewing a health plan in Alabama — including third-party contractors performing utilization review on the payer's behalf.

This is not an anti-AI bill. It's a liability assignment bill. The distinction matters for how RCM teams should read it.

What the Bill Actually Requires

Four provisions are worth understanding in detail:

• Individual-level decision making. AI tools cannot rely solely on group-level data to make prior authorization determinations. Decisions must be based on the member's specific medical history, the clinical circumstances presented by the provider, and additional clinical information in the medical record. This directly targets the practice of using population-level utilization patterns as a proxy for individual medical necessity.
• Physician sign-off on denials. Any decision to deny a prior authorization request must be made by a physician or other qualified provider — not by the AI system alone. The AI can inform the review. It cannot own the outcome.
• Annual non-discrimination certification. Payers must certify to the Alabama Department of Insurance each year that their AI systems don't discriminate against any subscriber group and produce consistent decisions for enrollees with similar circumstances.
• Transparency at the point of decision. Written notice must be provided to group plan sponsors and individual members whenever AI is used as a tool in utilization review.

Why This Is a Bigger Deal Than One State

Alabama isn't acting in isolation. A growing number of states — including Colorado, California, and Texas — have moved to regulate how payers use AI in coverage decisions over the past two years. The federal RAISE Act (reintroduced this week at the Senate level) targets similar behavior in Medicare Advantage. What's happening at the state level is the regulatory floor rising.

The enforcement mechanism in Alabama is also meaningful. The state's insurance department would be authorized to investigate payers believed to be making adverse PA decisions without considering individual medical history. First violations trigger required corrections. Repeat violations carry disciplinary consequences. That's not a reporting requirement — that's an enforcement structure.

What to Watch For

The bill still has to clear the House Insurance Committee. A few things worth tracking:

• House committee timeline. Alabama's legislative session runs through June. Given the Senate vote was 29-0 (unanimous), floor opposition is unlikely — but committee rewrites happen.
• The "qualified provider" definition. The bill requires that denials be made by "a physician or other qualified provider." How broadly that second category gets defined matters for payer compliance cost and which roles can be used for the sign-off requirement.
• Reciprocal pressure on other state legislatures. If Alabama passes this, expect similar language to appear in legislative sessions across the South over the next 12 months. The template is now written.

For providers and billing teams operating in Alabama, the practical action is documentation. If you have a pending denial pipeline where AI was involved and the clinical specifics weren't adequately documented in the request, now is the time to build that habit. Under this law, a payer that relied on group-level data to deny a claim would be in violation — and your documentation is what substantiates that argument in an audit or appeal.