Group CUO of Zurich on the evolution of underwriting

How will GenAI impact the role of the underwriter going forward?

Group CUO of Zurich on the evolution of underwriting


By Mia Wallace

As a firm believer in the transformative power of underwriting, Penny Seach (pictured), group chief underwriting officer at Zurich Insurance, is happy to plant the flag that underwriting “is the heart and soul of an insurance organization”. There are few in the industry quite as well placed to make that declaration.

How the role of the underwriter has – and hasn’t – changed

Discussing the future of underwriting in a recent interview with Insurance Business, Seach explored how underwriting has – and hasn’t – changed in the course of her career to date.

“What has changed is the fast-moving complexity of the world around us and what we do,” she said. “Even from a year ago, there is now more complexity. So, from an underwriting perspective, how we’re thinking about risk, understanding and digesting it continues to evolve at a very rapid pace.

“In terms of underwriting as a career, using critical thinking skills to make an assessment, and then going through the process of underwriting and determining the right capacity and the right terms and conditions for a risk has always been key. That, in essence, hasn't changed but the complexity of the changing world with the rapid evolution of technology offers an environment in which an underwriter can thrive.”

The impact of new technologies on underwriting and underwriters

Seach highlighted how new and innovative technologies, including generative AI (GenAI), are being used to complement the skills underwriters bring to the table and to create new speed and scale of decision-making. When she thinks about what keeps her up at night, she said, it’s that intersection of concentration risk and concentration exposure.

The insurance market is facing numerous complex risks and the increasing connectivity of these risks, alongside the rapid evolution of technology, she said.  From an underwriting perspective, it’s critical to consider what that intersection might look like and start to find ways to leverage it for the benefit of insurance businesses and, importantly, the customers they serve.

“The other interesting bit is, traditionally, we’ve used historic data to be predictive of what we think future trends are going to be,” she said. “Clearly, there’s still a place for that but you need to have an overlay of those changes as well. Relying solely on historical data, you may go off the wrong path; you need to take into account how things have changed, and where they might be headed next.

“There is more uncertainty when considering forward-looking trends, but I think that we can use technology in a smarter way to help us, not to make decisions for us. We use scenarios to understand concentration risk and what that could mean for us. We can and do use AI at pace now to help us model what those scenarios could be, and gain new insights into them. Previously, you’d input the problem in and get the modelled results, whereas now, you can understand the pathway of that thinking.”

This understanding allows underwriters to think more creatively about risk, Seach said, and to explore different scenarios which branch away from the original problem. This is one of the key ways in which she is seeing first-hand how technology enables underwriters to engage and interact more thoughtfully with potential areas of systemic exposure or concentration risk.

Balancing short-term and long-term threats and challenges

As somebody who has spent a lot of her career involved in liability underwriting – which is long-tail underwriting, Seach understands the delicate task of balancing real-time threats with long-term exposures.

“Every underwriter has their own perspective, of course, but as liability underwriters, we always say we’ve got the most interesting role,” she said. “Because we think about today based on what happened yesterday and what it means for tomorrow… I think bringing AI to that problem statement, helps you to be more thoughtful about what that intersection looks like.

“It’s easy to become overwhelmed when you’re thinking through different risk scenarios. But using AI to support your thinking can be empowering because it helps you visualize potential risks and how to manage them in different ways.”

Understanding and harnessing the full potential of GenAI

The power of GenAI, and LLMs, such as ChatGPT, from an underwriter’s perspective is less about how you engage with your data, she said, and more about training your brain to ask the right questions.

Seach noted that, from a Zurich perspective, the ability to access and perform advanced analytics at pace and at depth supports the underwriting process rather than replacing it. The goal is to access information and provide new insights to underwriters so they can engage with the data and ask the right questions.

“This helps us to understand our customers in a more intimate way and that, from our perspective, is what really makes the difference,” she said. “After all, insurance is and will always be about relationships. It’s about trust, about credibility, and about how we interact with our customers.”

Related Stories

Keep up with the latest news and events

Join our mailing list, it’s free!