Three States Pass Laws Barring AI-Only Health Insurance Claim Denials

--- headline: "Three States Pass Laws Barring AI-Only Health Insurance Claim Denials" slug: state-ai-health-insurer-laws-2026 category: policy story_number: "13" date: 2026-04-27 authors: - The Vault AI Staff sources: - name: Alston & Bird AI Quarterly url: https://www.alston.com/en/insights/publications/2026/04/ai-quarterly-april-2026 - name: National Law Review url: https://natlawreview.com/article/state-legislatures-consider-oversight-artificial-intelligence-health-insurance - name: KFF Health News url: https://kffhealthnews.org/news/article/artificial-intelligence-ai-health-insurance-companies-state-regulation-trump/ - name: Governing url: https://www.governing.com/artificial-intelligence/red-and-blue-states-want-to-regulate-ai-in-insurance-the-white-house-disagrees ---

# Three States Pass Laws Barring AI-Only Health Insurance Claim Denials

When a health insurer denies a claim, patients expect a human being looked at their case. In Indiana, Utah, and Washington, that expectation is now the law.

All three states enacted legislation in early 2026 that restricts health insurers from using artificial intelligence as the sole basis for denying, modifying, or downcoding medical claims. The new laws mark the most concrete legislative response yet to a growing national outcry over algorithmic decision-making in healthcare, and they arrive as state legislatures across the country grapple with more than 600 AI-related bills this session alone.

What the Laws Require

Indiana moved first. Governor Eric Holcomb signed House Bill 1271 on March 4, 2026, with an effective date of July 1. The law prohibits insurers from using any automated process, system, or tool -- including artificial intelligence -- as the sole basis to downcode a claim based on medical necessity unless a human employee or contractor has first reviewed the covered individual’s medical record. The bill also imposes a parallel obligation on providers, barring them from using AI to submit claims without human review.

“This is a commonsense guardrail,” said Indiana State Representative Martin Carbaugh, chair of the House Insurance Committee and a sponsor of HB 1271. “We are not banning technology. We are saying that when it comes to a decision about someone’s health care, a human being has to be in the loop.”

Utah followed on March 19, when Governor Spencer Cox signed Senate Bill 319, the Health Insurance Preauthorization Amendments Act, effective January 1, 2027. The law requires insurers to publicly disclose whether they use artificial intelligence -- including generative AI -- in reviewing authorization requests. Insurers must notify the Utah Department of Insurance, every in-network provider, and each enrollee if AI plays a role in preauthorization decisions. The bill also establishes firm decision timelines: seven calendar days for standard requests and 72 hours for urgent cases.

Washington rounded out the trio with Senate Bill 5395, which targets prior authorization practices directly. The bill prohibits the use of AI for prior authorization denials, requires easily accessible notices of prior authorization changes, and bans retrospective denials -- a practice in which insurers retroactively reverse coverage decisions after care has already been delivered.

A Bipartisan Groundswell

What makes these three laws notable is not just their substance but their political geography. Indiana and Utah are solidly Republican-governed states; Washington is led by Democrats. The bipartisan nature of the push reflects polling data showing that concern over AI in healthcare transcends party lines. A December 2025 Fox News poll found that 63 percent of voters describe themselves as “very” or “extremely” concerned about artificial intelligence, with nearly two-thirds of Democrats and just over three in five Republicans expressing worry.

“This is not a red-versus-blue issue,” said Matt Fisher, a health care regulatory attorney at Cinnamon Health Advisors and a contributor to the National Law Review. “When voters hear that an algorithm denied grandma’s hip replacement without anyone looking at her chart, they don’t care which party fixes it. They just want it fixed.”

The state-level momentum is significant because it is filling a federal vacuum. The Trump administration’s December 2025 executive order on AI sought to preempt most state regulation, describing “a race with adversaries for supremacy” and threatening to withhold federal funding from states that enact what it characterized as “excessive” AI rules. But health insurance regulation has historically been a state-level function, and legislators in both parties have shown little appetite for ceding that ground.

The Bigger Picture

The three enacted laws are the leading edge of a much larger wave. According to the Manatt Health AI Policy Tracker, 43 states have introduced over 240 bills related to AI in health care in the first months of 2026 -- nearly as many as were introduced in all of 2025. Bills with downcoding provisions alone have appeared in seven states, including California, Connecticut, Illinois, Maryland, Missouri, and Oregon in addition to Indiana.

Georgia’s Senate unanimously passed Senate Bill 444, which would prohibit coverage decisions from relying solely on AI systems. Minnesota’s HF 2500 would bar carriers from using AI to make adverse prior authorization determinations. Even Florida, where Governor Ron DeSantis has taken a skeptical view of broad AI regulation, saw Senate Bill 202 introduced to require human sign-off on claim denials, though the bill stalled in the Banking and Insurance Committee.

The insurance industry has not been uniformly opposed. Trade groups have generally accepted the principle of human oversight while pushing back on prescriptive technical mandates and arguing that AI-assisted claims processing actually reduces errors and speeds approvals for the vast majority of routine claims.

What Comes Next

The tension between state action and federal preemption is likely to intensify as more bills advance through legislatures. The National Association of Insurance Commissioners addressed the issue at its 2026 Spring National Meeting in San Diego, where working groups on Big Data and Artificial Intelligence discussed frameworks for consistent state-level oversight without triggering a federal preemption fight.

For patients, the immediate takeaway is straightforward: in Indiana, Utah, and Washington, an algorithm alone can no longer stand between them and their coverage. Whether the remaining 47 states follow -- and whether the White House tries to stop them -- will be one of the defining policy battles of the year.