North Carolina senators push AI guardrails for insurance claims and medical billing
North Carolina lawmakers are considering restrictions on how insurers and hospitals use artificial intelligence to deny claims and select billing codes, citing evidence that the technology inflates healthcare costs without human review.
Senator Amy Galey (R-Alamance) introduced a committee substitute to House Bill 565 on Wednesday that would require a human to review any insurance claim denied by AI. "If the AI does not approve the claim, it must be reviewed by a living human being," Galey told the Senate Health Committee. "We do not want robots or computers to deny care to sick people."
The bill also targets hospital billing practices. Galey said hospitals are using ambient AI-software that listens to doctor-patient conversations-not just to streamline medical records but to select the highest-paying diagnosis codes for a given condition.
"There is evidence that AI and hospital billing is driving the appearance of more serious diagnoses without corresponding changes in treatment," Galey said. She cited a Rice University Baker Institute report showing hospital service prices have surged more than 200% since 2000.
Galey argued the cost increases stem from billing practices, not from sicker patients or higher labor costs. The more serious a diagnosis code, the more a hospital can charge.
Opposition and concerns from healthcare providers
Sen. Gale Adcock (D-Wake), a family nurse practitioner, questioned whether new restrictions were necessary. Upcoding-selecting a higher billing code than warranted-is already illegal, and the state Department of Health and Human Services can already suspend or terminate a provider's license for the practice.
Adcock called the bill "one-sided" and said it failed to address differences in charges between inpatient and outpatient services. "I think this bill needs a great deal of work on it to achieve the balance we're looking for, and also the ability to enforce it," she said.
Sen. Natalie Murdock (D-Durham) said the two largest hospitals in her district worried about how the legislation defines upcoding and fraud. Hospitals feared being labeled fraudulent if they selected a code they believed was clinically appropriate.
Murdock called for more input from healthcare organizations. "As a state, we haven't even scratched the surface with what other states are doing in the AI space, but medical billing and coding is going to be a real issue. I just want to make sure we get it right," she said.
The North Carolina Healthcare Association said AI coding and billing are complex and said the group looks forward to working with lawmakers to develop appropriate regulations.
A second committee hearing scheduled for Thursday was canceled.
Learn more: AI for Healthcare and AI for Insurance cover how the technology is being deployed in these sectors.
Your membership also unlocks: