Operationalizing AI Fairness in Healthcare: A Case Study From Humana

As a leading payer, Humana’s top priority is making sure that its 22 million members have the best possible care and health outcomes, and it’s leaning on AI to help it accomplish that mission. Central to that challenge is ensuring health equity, particularly as it pertains to levels of care across underrepresented and minority groups.

Humana has developed a set of Responsible AI best practices to help it reduce the risk of potential harm from the 300+ models that it has running in production. I will talk about how my team and I have implemented end-to-end model monitoring for algorithmic bias and performance issues with the help of Arthur, the leading machine learning model monitoring platform.