Insurance partner and leader of Faegre Drinker’s AI-X team Scott Kosnoff spoke with American Banker about Colorado’s amended AI Act, which applies to automated decision-making technology that materially influences "consequential decisions" including approvals for loans and new accounts.
American Banker reports that the law requires financial institutions to provide consumers with instructions for requesting additional information when an AI model materially influences a loan or new account denial, including identifying every data source used in the decision. Consumers may also access and correct the personal data used and obtain "meaningful" human review.
Kosnoff emphasized that the law’s granularity is particularly noteworthy, citing the requirement to identify every data source by name. He added that the data correction and meaningful human review provisions may require institutions to build new infrastructure.
“Banks should invest in an AI governance framework that gives them a good story to tell - one that demonstrates they understand AI-related risks and have taken reasonable steps to mitigate them," Kosnoff said. “These systems can act autonomously with limited human oversight, creating risks of unauthorized actions and data exposure that can occur at scale.”