Connecticut has new regulations for the use of artificial intelligence in the State Employee Health Plan and Partnership Plan, which serve more than 270,000 members.
This comes in response to growing warnings about the unchecked growth of AI in the health insurance industry, State Comptroller Sean Scanlon said at the announcement of the regulations on Wednesday.
“In my mind there is a difference between being able to use ChatGPT to do your homework and somebody making health care decisions for you. And we have to be very, very clear about that,” Scanlon said, who oversees the state health plans.
The regulations require a human review of all adverse healthcare decisions.
“Carriers can’t use AI or predictive models as the sole basis to down-code claims, reduce provider payment, or alter billing codes without a human review,” Scanlon said.
The regulations mandate protecting patient privacy, require person-to-person communication, and require transparency audits of AI system accuracy.
And carrier governance and audit procedures must be disclosed to the comptroller.
