Legislative Moves on AI Oversight in Insurance Claims

Investigations have found that major insurers are using algorithms to generate denial letters in bulk. A Florida House panel has unanimously approved a bill that would require a human to make the final decision on whether an insurance claim is denied or reduced.

The proposal aims to create a safeguard against the increasing use of artificial intelligence in the insurance industry, ensuring that automated systems alone cannot reject claims. Representative Hillary Cassel, the bill's sponsor, stated, "No Floridian should ever have a claim denied based solely on an automated output." The House Insurance & Banking Subcommittee passed the measure on Tuesday, gathering support from hospital and doctor associations while facing opposition from insurance industry groups.

This legislative move comes during what House Speaker Daniel Perez has declared Artificial Intelligence Week, a period dedicated to examining the impact of rapidly advancing technology. The bill clarifies that while insurers can use AI and algorithms to process claims and recommend approvals or denials, the ultimate authority must rest with a qualified human professional.

The decision to deny a claim or any portion of a claim must be made by a person, according to the bill. The push for human oversight arises amid growing concerns over high claim denial rates. In 2023, data from KFF showed that 20 percent of all claims under Affordable Care Act marketplace plans were denied.

Similarly, audits revealed that insurers frequently deny or delay millions of care requests within Medicare Advantage, which covers more than 30 million Americans. Despite this, fewer than 1 percent of ACA plan denials are appealed.

Florida residents have also complained about high denial rates on claims for damage caused by hurricanes and other major storms, which have been exacerbated by AI. Investigations revealed that major insurers used algorithms to generate denial letters almost instantaneously and against the recommendations of treating physicians.

A 2022 ProPublica report showed that Cigna used an AI-driven system to deny more than 30,000 claims in just two months, often without a human review. In the following year, a class-action lawsuit accused UnitedHealth of using its AI tool, nH Predict, to prematurely cut off care for elderly patients.

This lawsuit was filed in 2023, just a year before UnitedHealth CEO Brian Thompson was murdered on a New York City street, prompting wide discussion about how insurance giants use denials to control costs.

Representative Cassel noted that the genesis of this bill stemmed from the murder of the United Healthcare CEO, highlighting that one of the alleged motives was the company's claim denial practices. She stated that there is currently a class action showing that 90 percent of their claims were denied with errors when they utilized AI.

Representatives from the Florida Insurance Council, the American Property Casualty Insurance Association, and the Personal Insurance Federation of Florida argued against the bill, claiming it could slow down the claims process and increase consumer costs, while asserting that existing laws already ensure fair claims handling.

The states insurance commissioner, Michael Yaworsky, expressed a desire for a regulatory path that ensures responsible use of AI in the industry. An amendment has expanded the bill to include workers compensation and HMOs, and the bill now heads to another committee before it can reach the House floor.