AI & Algorithms: Reshaping portfolio checkups — and outsmarting risk

For most people, “algorithm” is a word they've heard only in the context of their social media feed. For tech-savvy insurance providers, however, AI-powered algorithms are taking on a major role in automating underwriting needs.

“Algorithms don’t have to be scary and they don’t have to be only focused on the simple stuff or follow business. They can be applied to messy, specialty, complex lead underwriting processes,” Bijal Patel, co-founder and CTO of MGA Aurora said during a recent INFUSE webinar hosted by Send Technology.

Her fellow panelists, Thomas Nasso, CUO, Falvey Insurance, and Dan Walsh, CUO, Ardonagh Specialty, explained how algorithmic underwriting can apply not just to individual products but also expand to look at a client’s entire portfolio.

“Underwriting has always been sort of a combination of art and science,” Nasso said. “I think we’re going to tip more towards science, more towards the direction of algorithms, and everybody should be utilizing it. But there is still that piece of the art in underwriting that I think will last throughout.”

“AI and predictive analytics have been used for many years, particularly within the pricing processes and also to analyze historical claims data. The advantage we have now is other use cases of AI and algorithms,” Patel said.

What is algorithmic underwriting?

Algorithmic underwriting is the use of AI algorithms to automate the underwriting process. In recent times, insurers have begun using different forms of AI for individual case underwriting. But the panelists noted that algorithms don’t have to stop there; they can also be useful for lead underwriting and portfolio underwriting.

Further, Patel emphasized that automating the manual steps an underwriter would take for risk assessment frees up their time to focus on:

• Building broker relationships, which helps with strategic growth
• More effectively managing underwriting performance and loss ratio performance at the portfolio level
• Using the predictive analytics on portfolio performance to find new risk factors that would improve pricing and underwriting

“Instead of underwriters having to rekey data or offshore teams having to rekey data, AI can be used to automatically extract that data and push it through that automated underwriting workflow,” Patel said.

For example, she said, if an insurer decides to stop underwriting properties with slate tile roofs, that rule could be applied consistently across the entire portfolio instead of having to rely on underwriters to apply it individually for each case.

Gaining traction

Algorithms are particularly gaining traction for portfolio underwriting, according to Nasso, and the impact of AI on the insurance industry itself is partly to blame.

“Portfolio underwriting and the use of data to drive underwriting is the result of the insurance market’s need to address the intensity for speed, turnaround time for the brokers and insureds as well as creating efficiencies within the organization for underwriters and portfolio managers to handle larger books of business while creating greater margin from an expense standpoint,” Nasso said.

He noted that AI-powered portfolio underwriting aims to provide:

• Improved decision-making based on appetites and structures laid out
• Enhanced risk insights to improve risk management of portfolios
• Quicker decisions in portfolio segmentation steering
• Improved pricing adequacy and profitability of the portfolio by identifying trends quickly

Patel added that return on investment can include efficiency through improved productivity, but also better effectiveness through writing better business and achieving a better claims ratio.

Barriers to adoption

However, the panelists all agreed that there are barriers to adoption of AI and algorithmic underwriting — some of which are somewhat endemic to the insurance industry:

1. Clean data
2. Legacy systems
3. Company/industry culture
4. Ease of use
5. Investment in technology

Clean, accessible data for AI tools to use is a “prerequisite” that many experts have found insurers tend to struggle with.

However, Patel said she sees this as an opportunity for insurers to use AI to help clean up that data and “start getting a better, more granular view, more consistent and more structured view of data.”

“I think a lot of insurers are hesitant about algorithms because they don’t have that clean data, but there is opportunity to start fresh,” she said. “So even if you don’t have this accessible, clean historical data, start collecting it now.”

Similarly, Nasso said many insurers have legacy systems that contain data which has not been mapped correctly for AI, holding carriers back.

“Legacy systems in the insurance world are what's preventing this from really taking off. I think that’s why we have insurtechs coming in, other people coming in,” he said.

Walsh added, however, that some barriers are not so much technical as they are about behavior — namely the insurance industry’s general AI hesitancy. He noted that some insurers are reluctant to invest in the technology needed for an AI-powered strategy, while in other cases, the tools being adopted must be easy for staff and brokers to use.

“We’ve got to invest in these tools because they’re very good,” Walsh said.

Aurora is an MGA offering algorithmic underwriting as a service. It was founded in 1957 and is based in New York.

Falvey Insurance Group focuses on specialty products such as marine cargo and shippers insurance and is based in Rhode Island.

Equinox is an in-house underwriting unit for the Ardonagh Group, a global specialty insurance broker headquartered in the UK.

Send Technology Solutions is an insurtech that provides commercial insurance software solutions through an online platform and is based in London.

© Entire contents copyright 2025 by InsuranceNewsNet.com Inc. All rights reserved. No part of this article may be reprinted without the expressed written consent from InsuranceNewsNet.com.

Rayne Morgan is a journalist, copywriter, and editor with over 10 years' combined experience in digital content and print media. You can reach her at [email protected].