A growing number of P&C insurers are discovering that the key to gaining the most value from investments in generative AI isn’t (just) identifying and implementing the most compelling use cases. It’s an enterprise-wide approach to implementing GenAI across the entire insurance value chain.
According to a study from EY-Parthenon, 73% of large carriers have already invested in generative artificial intelligence, while an additional 17% plan to soon. For more than half of them, key drivers for GenAI adoption include productivity enhancements, cost reductions, and new revenue generation. Two-thirds anticipate a revenue lift of greater than 10% from GenAI implementations within core functions.
It’s easy to see why. Among other things, GenAI can dramatically enhance the usability of AI and analytics applications through natural language interfaces and by synthesizing information and making it easily consumable for workers. But after the rush of enthusiasm for generative AI over the past 24 months, many carriers are stuck in what McKinsey calls “pilot purgatory.”
While crucial, trials in isolated pockets of claims management, underwriting, distribution, fraud reduction, or customer self-service can only take you so far for so long. Ultimately, a disjointed approach to GenAI can lead to duplicative efforts, disconnected models, and no clear path to the scalability needed to achieve substantive savings or revenue generation. In other words, launching GenAI pilots is one thing. Scaling them to create meaningful ROI is entirely another. But it doesn’t need to be this way.
GenAI 1.0: Disjointed Efforts, Fragmented Results
As the EY-Parthenon report points out, most carriers are prioritizing GenAI use cases to transform a specific part of the value chain, with an emphasis on quick wins.
To be sure, the pilots seen within our industry are incredibly promising. Among the most common are self-service, and extracting insights from unstructured and semi-structured data. In underwriting, this can mean making it easy to pull information from policy submissions and weighing them against the carrier’s risk appetite and underwriting guidelines to accelerate time-to-quote. Together with new distribution models, Bain & Company estimates these GenAI-enabled productivity gains could help carriers increase revenues by up to 20% while reducing costs by up to 15%. In new product development, generative AI may be leveraged to identify opportunities to meet evolving customer needs.
Meanwhile, GenAI-based tools that offer claims summaries, next-best-actions, and procedural guidance have been shown to improve an agent’s claim handling times by 60%. By Bain’s estimates, this could help reduce loss-adjusting expenses by up to 25%, and leakage by up to 50%—creating more than $100 billion in benefits for insurers and customers. According to one January 2025 report, some industry observers suspect this technology is already generating enough savings to help some lines improve profitability despite rising costs.
But increasingly, insurers are looking at the benefits produced by their first-generation GenAI use cases and asking what it’ll take to scale them into the fabric of the entire organization? How do we extend them across different domains, business models, and geographies?
With these questions in mind, a growing number of organizations are shifting their GenAI mobilization efforts from a Gen AI 1.0 model—strictly bottom-up, use case-centric experimentation—to something you could call GenAI 2.0, a more centralized, enterprise-wide approach with direct ties to the C-suite. According to McKinsey, scaling generative AI for maximum ROI requires a comprehensive approach that combines GenAI with more-traditional forms of AI and robotic process automation in order to rethink processes across all domains.
Going All-In with Generative AI 2.0
This kind of centralized, enterprise-wide approach includes establishing a long-term strategic vision, driven by the CEO and executive team, to ensure generative AI efforts are aligned with overall business goals and receive the funding needed to succeed.
In EY-Parthenon’s 2024 survey, 41% of P&C insurers are currently using this kind of top-down model. And 57% of insurers that have yet to launch but are planning to establish a dedicated GenAI team expect to leverage this more centralized approach. It’s worth noting that another 41% embrace a dual track approach that balances top-down strategies with grassroots experimentation—a model EY believes is most advantageous.
Whether they pursue a 100% top-down or dual-track model, carriers operating a modern, cloud-based infrastructure will have an advantage. Using Guidewire as an example, our vision for Generative AI 2.0 is a platform-wide approach to integrating generative AI capabilities across all of a carrier’s core operations.
Among other things, this would simplify back-end data management and systemwide governance and expedite product innovation. Generative AI applications would have instant access to all carrier data without creating added latency or requiring additional compute power. And it would ensure consistency and accuracy of GenAI-enabled insights across every customer and agent touchpoint. What’s more, it would make it possible to leverage reusable components, which can accelerate development of new GenAI use cases by 30% to 50%.
A Platform Approach to the Next Phase of the GenAI Revolution
Early adopters of enterprise-wide AI bear out the benefits of this approach. According to one study, insurers leading in leveraging advanced forms of AI at scale achieve financial performance that’s 6X that of industry laggards. Check out the video below to learn more about our vision for platform-wide GenAI adoption, and how it’s designed to empower carriers to unlock the full potential of generative Ai throughout the insurance lifecycle.
