How insurers use AI to judge risks

David Gritz, Greg Hendrick and Gaurav Vasisht, seated in panel discussion.
David Gritz, co-founder, InsurTech NY; Greg Hendrick, CEO, Vantage Group Holdings; and Gaurav Vasisht, CEO and executive director of the New York State Insurance Fund (NYSIF), speak at InsurTech NY on March 31, 2026.

Takeaways:

Processing Content
  • Good governance supports efficiency
  • To use AI, insurers should first sort simple and complex risks
  • AI can improve insurers' management of staff

Insurance carriers are finding AI effective for governance, risk reduction and personnel management, according to executives who spoke at InsurTech NY.

David Zweier of Ascot Group
David Zweier, senior vice president of IT strategy and transformation at Ascot Group.
LinkedIn

"People can get scared of governance. They hear governance and think that [it] doesn't compute with AI, because it will slow me down," said David Zweier, senior vice president of IT strategy and transformation at Ascot Group, at the event in late March. "But effective governance actually enables agility, flexibility and speed if it's done properly."

Business and IT stakeholders must work together on AI and should be empowered to make decisions, Zweier added.

Patrick Gallic of Tokio Marine
Patrick Gallic, head of the international AI hub at Tokio Marine.
LinkedIn

Just adding AI by itself is not enough to make a difference, according to Patrick Gallic, head of the international AI hub at Tokio Marine. 

"The question then becomes, what is the benefit to the organization?" he said at the event. "If you're just deploying ChatGPT Enterprise, or Claude Cowork, and you're expecting to transform your business, that's not realistic. It's going to get adoption, but you are not going to transform the way that you work."

Wil Lombardi of GBU Life
Wil Lombardi, chief operating officer at GBU Life.
LinkedIn

Using an API gateway can ease integration of AI into an insurer's systems, as Wil Lombardi, chief operating officer at GBU Life, said. Before doing so, insurers must look at where they need to reduce risk, he added. There may be easy pieces of data to process, such as changes in address or beneficiaries. 

"If you think of policy services, we look at what are the high-volume, less risky ones to change or enhance," Lombardi said at the event.

Adam Landau of Berkshire Hathaway
Adam Landau, chief information officer at Berkshire Hathaway Guard.
LinkedIn

Adam Landau, chief information officer at Berkshire Hathaway Guard, a P&C insurer, said that while carriers can set criteria for acceptable risks and then use AI to process coverage applications that fit that criteria, more complex risks require human underwriters' attention.

"Don't do it just to put in a button, because somebody said you have to put in a button," Landau said at the event. "Do it for assessing risk, understand what your tolerance is, and then put the human in the loop with that."

The responsibility for risk decisions rests with carriers, not their service providers, Landau stressed. However, AI can free underwriters to pay more attention to such complex risks, he explained. 

"It's really about now freeing up skilled workers, underwriters, claims representatives, whoever it is, to start working on the higher-value stuff — stuff that they would have glanced over, maybe not had enough time to deal with — to think about a complex risk that an agent said wasn't the best risk to write," Landau said.

Gaurav Vasisht - NYSIF.jpg
Gaurav Vasisht, executive director and CEO, NYSIF.

The New York State Insurance Fund (NYSIF), a workers comp insurer that had developed its own mobile technology to serve policyholders, is considering the use of AI for predictive modeling and processing claims, according to Gaurav Vasisht, its CEO and executive director. He sees potential from using AI, without replacing staff.

Generative AI "makes the job of being a claims manager much more rewarding," he said at the event. "It frees up people's time, and gives them some level of assurance that they're not going to be laid off tomorrow. If we stay true to this, it will pay dividends over the medium and long term."

Related stories:

For insurers, AI hype ends and decision-making begins

Understanding the risks of AI integration across the insurance industry

AI brings severe change but will "unlock" value for insurance


For reprint and licensing requests for this article, click here.
Artificial intelligence Insurtech Property and casualty insurance Life insurance Commercial insurance
MORE FROM DIGITAL INSURANCE
Load More