A bipartisan group of House lawmakers say they will introduce legislation to set guardrails for the use of artificial intelligence in health care.

The lawmakers say the bill will provide new regulations for the way AI is used by insurance companies, hospitals and health care providers. The proposal comes as AI is increasingly being employed in a variety of ways, including by insurance companies in determining whether to cover treatments.

“Getting an unjust denial from your insurance company is bad enough, but knowing that AI and not humans are making the decisions is outrageous,” said Rep. Tarik Khan, D-Philadelphia.

An American Medical Association survey released earlier this year found half of the doctors surveyed said they are most concerned about the use of AI by health insurance companies.

“These AI tools have been accused of producing high rates of care denial — in some cases, 16 times higher than is typical,” the AMA said, citing a congressional report released last year.

At the same time, the percentage of doctors who reported using AI in their practices jumped from 2023 to 2024. Nearly two-thirds of physicians, 66%, surveyed reported using AI in health care last year, a 78% increase over the prior year, according to the AMA survey.

The doctors reported using AI for billing codes, with visit notes, to create discharge instructions, for translation and to assist with diagnoses.

To prevent clinicians from overly relying on AI, the legislation would also ensure that a human decisionmaker would make the ultimate decision based on an individualized assessment when AI is used by insurers, hospitals or clinicians.

“As AI becomes a bigger part of our day-to-day lives, we must make sure it is not being over-relied on in our health care system,” said Rep. Joe Hogan, R-Bucks. “This legislation would make sure that there is still a human element in place when determining life and death decisions.”

The legislation would also require an attestation by insurers to the PA Department of Insurance and hospitals and clinicians to the PA Department of Health that bias and discrimination already prohibited by state law have been minimized in their usage of AI and the providing of evidence of how that determination was made.

The legislation has not yet been formally introduced.