Artificial intelligence and machine learning are enablers, making space for humans to do more important things, like close the insurance protection gap.

That was the word from Natanje Holt, DXC Technology, United Kingdom and Ireland insurance software general manager, speaking in a panel at the Bermuda Risk Summit 2026.

She said it was about taking some of the heavy lifting out of labour so employees can focus on more important things such as client relationship building or innovation.

“The insurance gap is enormous,” she said. “There is so much more we can do.”

Ms Holt was excited to see how the future of work evolves and predict it would bring many more jobs.

However, when questioned she was unable to specify where those jobs would be.

Her firm, DXC is one of the world’s largest insurance software providers with over 2,000 customers.

Ms Holt broke her clients’ AI use into three categories: analysing and gathering data, making decisions and acting.

“There are different maturity levels across the three,” she said. “Analysing and gathering data is fairly mature by our lens.”

She saw less appetite among clients to use AI to make or help with decisions.

In terms of acting, AI had potential for automating processes such as changing an address, but was still very much assisting, she said.

“In terms of acting and decision-making AI still has a ways to go to mature,” she said.

Another panellist, Andrew Johnston, global head of Insurtech Gallagher Re, reported seeing the most impact from artificial intelligence in sniffing out claims fraud.

Speaking at the Hamilton Princess and Beach Club, he said it was interesting to see the fraud-detecting AI technology at work.

Mr Johnston said AI often not only affirms employee suspicions about a transaction but also flags other things they may have missed.

“So the system is learning, and it is looking for affirmation in the loop,” Mr Johnston said.

David Flett, senior manager, IT advisory, KPMG Bermuda, warned that AI needs guardrails.

“There are some good examples already of things starting to go wrong for insurers who have not embedded good governance frameworks and good risk and control frameworks,” he said.

Henry Paddison, Relm Insurance chief operating officer saw a place for both AI and automation, depending on the goal.

“We are outcomes focused,” Mr Paddison said. “We are trying to make everyone who works at the company be more efficient, deliver things faster and in a smarter way.

“When we think about the uses of AI, we go back to that risk framework.”

When insuring a company that uses AI, Relm looks at the potential for physical harm, reputational risk, regulatory risk and financial risk.

“People always have to be involved to moderate that risk,” Mr Paddison said. “We see it as a tool, not some end-to-end function that is going to start making underwriting claims decisions for us. It is just supposed to help our workforce.”