Lawmakers in California are working to join other states regulating health insurers’ use of artificial intelligence tools in coverage decisions.
At least 40 states have introduced or passed legislation on AI regulation in 2024, with a half-dozen measures related specifically to the health-care industry, according to the National Conference of State Legislatures. A common focus of the health-specific bills has been greater oversight on insurers’ use of AI tools to expedite coverage decisions.
The effort comes as top insurers
California legislation passed with overwhelming bipartisan support by the Senate and set to go before a state Assembly hearing July 2 would require any health insurer in the state to have a licensed physician supervise the use of AI tools in the utilization review process.
Bills in New York and Pennsylvania would mandate insurers disclose to providers and individuals when AI is used. At least 11 states have also issued broad guidance on AI standards in health insurance based on a model released in December by the National Association of Insurance Commissioners.
Although state efforts to rein in AI are proliferating, they face constraints: it’s hard for states to change AI systems insurers use globally, and states don’t have authority over employer health plans subject to the federal Employee Retirement Income Security Act of 1974, or ERISA.
But legislators and health law analysts said in interviews that even if the federal government does more to establish holistic standards on how insurers use AI, states still have an integral role as the primary regulators of insurance.
As many state legislative sessions wrap up for the year, the introduction of legislation is bringing attention to the issue across the country, prompting pushback from insurer groups who argue the bills would lead to unnecessary overregulation of coverage decisions.
Lawmakers say the proposals can set off a wave of state actions that could force insurance companies to provide more information on their AI tools and how they are used to determine health coverage.
“These are essentially tools used to make life-changing, altering decisions for people,” said New York Assemblymember Pamela Hunter (D), the sponsor of the legislation there and vice president of the National Council of Insurance Legislators.
“We can’t wait—consumers need to be protected now,” Hunter said.
‘Set a Precedent’
Health insurance companies have been using AI tools for some time to help in “identifying gaps in evidence-based care to share with providers, improving consumer experience, speeding claims processes, and detecting fraud,” trade association America’s Health Insurance Plans wrote in July 2023 in response to the Biden administration request for information on national priorities for AI.
But the recent lawsuits against health insurance giants have brought forth allegations that AI tools have overridden doctors’ recommendations on needed care, forcing many patients to pay for treatment on their own or go without care.
A September 2023 study by KFF found one in five insured adults surveyed experienced a denied claim in the past year. Of those who experienced a denied claim, 55% said they ended up paying more than they expected for treatment or services.
“When we saw that there were lawsuits and accusations and concerns that insurance companies were algorithmically denying claims without physician oversight, that’s when we started working on this specific bill,” said California state Sen. Josh Becker (D).
Becker’s bill (SB 1120) passed the state Senate in May in a 37-0 vote with three abstentions. It would apply to both commercial and state-run health plans “to ensure that a denial, a delay, or modification of health-care services based on medical necessity be made by a licensed physician or other health-care provider competent to evaluate the specific clinical issues of the case,” Becker said.
Becker said he believes his legislation has the potential to “set a precedent and be picked up” by other states.
In Pennsylvania, where the 2024 legislative session lasts through November, state Rep. Arvind Venkat’s (D) bill (HB 1663) would require state-regulated health plans to provide information on the AI tools they use in their utilization review process to the state Department of Insurance. They would also be required to disclose the use of AI to health-care providers and patients.
Venkat, an emergency physician, said it’s likely the legislation will get a hearing in the fall in anticipation of a vote in the next legislative session. He noted, though, that his colleagues have questioned whether the legislation should be expanded to address the use of AI across the health-care industry.
“Should it be really a global regulatory effort related to the health care system as a whole and its use of artificial intelligence? I think that’s a fair point,” Venkat said.
Hunter said her New York bill (A 9149) will have to be reintroduced next session since the state’s 2024 session ended earlier this month.
Similar bills were introduced this year in Georgia and Oklahoma, though legislative sessions concluded without floor votes on the legislation.
In Oklahoma, Republican state Rep. Daniel Pae’s bill would require insurers to publicly disclose the use of AI in utilization reviews, as well as submit the algorithms and training data sets to the Oklahoma Insurance Department.
“AI doesn’t discriminate based on one’s political affiliation,” Pae said, adding there’s “really strong bipartisan support” for regulating this technology.
Pae, whose legislation advanced along with other AI bills out of the Oklahoma House Government Modernization and Technology Committee in February, said states are in the beginning stages of addressing how health insurers and others use AI and that the issue is likely to gain more momentum in the next year.
Patchwork of Rules
Insurer groups argue the state proposals would go too far in regulating their decision making because the limit of states’ authority could complicate establishing common standards on the issue.
AHIP, along with the California Association of Health Plans and the Association of California Life and Health Insurance Companies, oppose the current version of SB 1120. They wrote in a May 16 letter to Becker that the legislation “would substantially limit who is allowed to conduct utilization reviews by requiring the reviewing provider be within the same or similar specialty as the requesting provider.”
“If the treating provider disagrees with the reviewing provider, current law already affords a process for the resolution of the dispute through independent medical review,” the groups wrote, adding the provision “is unnecessary and will only add more costs to the health care system.”
There’s also the issue of regulating AI within a state’s borders.
While consumer protection and insurance laws have primarily been a state function, AI “doesn’t seem like an issue that should be regulated state by state, because the same stuff is being promulgated nationwide or even globally,” said Michelle Mello, a professor of law and health policy at Stanford University.
This leads to “a patchwork of different rules” health insurers may have to comply with across states, Mello said.
There’s some signs of movement at the federal level, as well.
A bipartisan group of nearly 50 members of Congress sent a letter Tuesday calling on the Centers for Medicare & Medicaid Services to clarify the agency’s February guidance that Medicare Advantage plans are required to make prior authorization determinations based on a specific individual’s circumstances, and not soley based on algorithms or software tools.
“We believe CMS must be more proactive in monitoring plans’ use of AI and algorithm-driven tools,” the lawmakers wrote, adding “plans continue to use AI tools to erroneously deny care and contradict provider assessment findings.”
