The use of artificial intelligence has expanded, and insurance companies are now using algorithms to quickly assess claims, according to a report by the nonprofit news organization ProPublica.
Typically, claims need to be individually assessed by a medical director, but the algorithms can bypass much of the process. ProPublica reports that Cigna, for example, denied more than 300,000 claims with only 1.2 seconds of review time in two months using this method.
Pennsylvania legislators, such as Rep. Arvind Venkat, D-Allegheny, are trying to ensure AI isn’t the final call.
“The human beings who are supposed to be responsible for this simply abdicated their responsibility,” Venkat said. “They just pressed click and, you know, the decision was made. And that’s not something that should be allowed, in my opinion.”
Venkat plans to introduce legislation to regulate the algorithms used for insurance claims.
The legislation would require healthcare professionals who review claims and use AI to document their review of records and have algorithms defined so they can be regulated and subjected to current laws.
Insurance companies would have to submit their AI algorithms to the Department of Insurance for review, in order to minimize the risk of bias.
Venkat says biases, whether they be over sex, race or other characteristics, are being confirmed with the new AI decision-making process. The biases have been observed by the National Institute of Health, which found insurance claims from women, non-White people and the unemployed are more likely to be denied.
Venkat, a physician, said he is not alone in his concern, citing an article from The Washington Post in which hospital workers said they hope for stronger regulation regarding AI, despite using it themselves.
“What I think is really, really important is that it is highly unusual for industry to say we need regulation in order to do this well before it goes out into society,” Venkat said.
He plans to formally introduce the legislation before the end of the month. It has bipartisan support from Reps. Tarik Khan, D-Philadelphia, Bridget Kosierowski, D-Lackawanna, Nick Pisciottano, D-Allegheny, Joe Hogan, R-Bucks and Rob Mercuri, R-Allegheny, who have signed on as co-sponsors.
“The advent of artificial intelligence has been momentous,” Mercuri said, “and as a legislator thinking about the well-being and safety of our society, I have a deep desire and keen interest to find opportunities for us to put guardrails around the use of AI. In particular, in areas where health, wellness and safety are concerned.”
Hogan said putting guardrails on AI is important given its increasing influence on everyday life.
“AI is likely to change just about every facet of our lives,” Hogan said. “I think it is prudent to join with my colleagues in a bipartisan manner and ensure that the technology is being used with certain guardrails to provide transparency to consumers and allow for prompt human review before a claim is denied.”
