(Key Takeaways from the Insurants ITI Europe Roundtable)
The commercial insurance sector is navigating a period of significant transformation, driven by the rapid advancement of technology and the persistent challenge of extracting value from vast quantities of unstructured data.
A recent roundtable discussion at Insurtech Insights London convened a diverse group of industry experts to explore these critical issues. The event, hosted at the Intercontinental Hotel, brought together representatives from major insurance carriers, global consulting firms and InsurTech companies, including data and analytics leaders, underwriting experts and innovation specialists. I did want to take a moment to thank the Roundtable speakers for sharing their experience and insight. It made for a vibrant discussion.
The Industry’s Data Conundrum
The roundtable discussion underscored the multifaceted nature of the challenges facing the industry, with a central theme being the issue of data accessibility.
Participants highlighted the vast amounts of valuable information and insight that remain trapped within a multitude of unstructured documents. Policies, binders, slips, Statements of Value (SOVs), loss runs, bordereaux and ACORD forms were identified as key examples of data sources that are often difficult to access and analyse effectively.
This “document dilemma” poses a significant hurdle for insurers seeking to streamline workflows and gain deeper insights.
Adding to this complexity is the proliferation of various initiatives aimed at addressing specific process, document and market needs. While these efforts demonstrate a commitment to innovation, they also contribute to the fragmentation of the technological landscape. Insurers are confronted with a myriad of emerging technologies and techniques, including Large Language Models (LLMs) and Small Language Models (SLMs), while the pace of technological evolution shows no signs of abating.
This fragmented environment, coupled with the high costs associated with technology ownership, creates substantial risks for insurers and significantly influences their decisions to build or buy technology solutions.
Two Paths: Tier 1 vs. Smaller Organisations
The roundtable discussion also illuminated the divergent paths often taken by Tier 1 firms and smaller organisations in their approach to AI implementation.
Tier 1 firms, with their extensive technology resources, scale, and maturity, often have the capacity to adopt a “mix and match” strategy, selectively combining various technologies to address specific needs.
Smaller organisations, however, typically require a more holistic approach, seeking comprehensive, end-to-end solutions rather than point solutions that address only isolated problems.
Generative AI. Promise and Practicalities
Generative AI (GenAI) is a key area of interest for insurers, offering the potential to transform various aspects of their operations. Participants discussed the ongoing GenAI pilots within the industry and the critical importance of carefully managing Total Cost of Ownership (TCO) to ensure that these initiatives deliver tangible cost benefits.
However, the roundtable also addressed the practical challenges of implementing GenAI. There was a consensus that directly loading extremely large documents, such as a 1000-page policy, into an LLM is generally not a viable approach. Effective orchestration of data processing and analysis is essential to achieve consistent and reliable results when working with complex insurance documents.
Furthermore, the creation of a reliable “ground truth” for validating AI-augmented results remains a significant hurdle. Overcoming this challenge necessitates the development of robust, embedded operating models for continuous validation and quality control.
Augmentation, Not Automation. A Guiding Principle.
A core theme that emerged from the roundtable discussion was the emphasis on augmentation rather than pure automation. Participants recognized that while AI offers immense potential to streamline processes and enhance efficiency, its greatest value lies in augmenting human capabilities.
The discussion highlighted the wide range of use cases for GenAI, spanning from supporting high-volume, industrialized processes to providing sophisticated tools for interrogatory and expert support and even for challenging existing expertise and biases.
Charting a Path Forward
The insights from the roundtable provide a valuable roadmap for insurers navigating the complexities of AI implementation. By acknowledging the multifaceted challenges, understanding the nuances of different implementation strategies and prioritizing augmentation, the industry can effectively harness the power of AI to unlock new levels of efficiency, insight and success.
