AI and E&O

Do you have concerns about E&O risk as technology evolves with AI?

“No. It’s super important for us to be paying attention to that as an industry, given the fiduciary responsibility that we have to our customers,” said James Thom, Vertafore’s chief product officer, adding that there’s two things that come to mind when it comes to this responsibility.

“One is in many cases when we’re releasing these new AI technologies, we’re very focused on accuracy and what’s happening in a way that it surpasses human accuracy that existed in the process previously. Two, we’re very focused on having a human in the loop in these processes so we get that human review. And three, we’re making sure that we’re tracking and recording the audit trail … what was done, why it was done, and where it got done … so that you can understand the order of operations that led maybe to a decision. Who was involved, what did the machine do, what did the person do along the way to reduce any risk associated.”

Thom said it’s important for agencies to ask these questions of their technology vendors and make sure to work with companies that are thinking about the E&O risk that they face and what does that actually mean for them and the exposure that they have.

Ross Sinclair, CEO at EIP, has a different point of view about the future risk of some AI.

“Autonomous AI presents an enormous challenge for insurers, particularly how they identify, model, and price this new risk,” Sinclair said in a statement. “Potential claims against executives add another layer of complexity, but identifying who might be held responsible does answer the question of which losses to the affected business are covered,” he said.

“The biggest concern for insurers is that some of this exposure may already be covered within conventional cyber policies that were neither designed nor priced with autonomous AI in mind,” he said. As a result, some insurers are making AI exclusions within cyber policies, or creating new AI-specific policies as a completely separate risk class outside of cyber.

Going forward, clarity on what is covered and what is excluded needs to be addressed, he added. “Insurers face a difficult enough task pricing an unpredictable risk without added uncertainty over what they have already agreed to insure.”

Topics InsurTech Data Driven Artificial Intelligence