Unlocking the Potential of Generative AI in Insurance
The 2025 Insurtech Insights Europe conference hosted a thought-provoking discussion on the tangible value generative AI (GenAI) can bring to general insurance today. Laura Drabik from Guidewire engaged in a dialogue with Tom Wilde, CEO of Indico Data, and Terry Buechner, the global insurance core systems lead at AWS, focusing on the practical applications of AI within the industry.
The Disruptive Power of GenAI
Wilde emphasized the significant disruption GenAI has brought, noting that last year, around 25 million individuals worldwide identified as software engineers. With the advent of GenAI, anyone with basic typing skills could essentially become a programmer, showcasing the profound impact AI has had.
“The challenge lies in managing this capability,” Wilde remarked. “Insurance companies wouldn’t suddenly allow everyone to alter pricing models.” The imperative now is for the industry to steer this technology responsibly, ensuring appropriate controls are established.
GenAI’s Strengths in Insurance
Wilde highlighted GenAI’s proficiency in summarization, particularly in summarizing underwriting and claims models with exceptional results. “For the first time, unstructured data, like underwriting guidance, can be transformed into a programmatic endpoint, enabling software and documents to communicate seamlessly – a true breakthrough,” he stated.
AI’s Role in Document Processing
Buechner shared insights on AI’s ability to process and summarize large volumes of documents. The much-discussed AI hype is warranted, especially in transforming document-heavy processes like claims and underwriting. He pointed out its significant role in claims management, from ID verification during the first notification of loss to achieving first-call resolution.
“When implemented effectively, GenAI can accelerate and enhance transparency in claims processing,” he asserted. While similar benefits extend to underwriting, complete automation, particularly in commercial lines, remains a complex challenge. “Human expertise and judgment are still crucial,” Buechner noted, “but the landscape is rapidly evolving.”
Challenges in Scaling AI Initiatives
Drabik referenced Deloitte’s findings that 76% of insurers have integrated generative AI in some capacity. However, only 15% have managed to scale these initiatives effectively. The pressing question is: what distinguishes those who scale successfully from those mired in endless pilot programs?
According to Buechner, a recent Boston Consulting Group study indicated only 26% of companies possess the capabilities to advance beyond proofs of concept and achieve real value. “This percentage is likely even lower in the insurance sector,” he acknowledged.
Strategies for Successful AI Scaling
To scale AI effectively, start with projects aligning with GenAI’s strengths. “Identify the problems you aim to solve or processes you wish to transform, then build from that foundation,” Buechner advised. Clear objectives and measurable success criteria are essential.
He emphasized cultivating a test-and-learn culture, a concept not traditionally ingrained in insurance. “This approach enables us to build on successful initiatives and retract when necessary due to technological or regulatory changes,” he concluded.
News
