Travelers Insurance rapidly deployed an AI-powered claims assistant nationwide, developed in close partnership with OpenAI, to enhance the first notice of loss process by improving customer experience, operational efficiency, and accuracy while maintaining human agent support. The company emphasizes responsible AI use, continuous improvement, and employee upskilling, with plans to expand AI applications across the claims lifecycle and other business areas.
Eric Rowan, Senior Vice President and Chief Information Officer at Travelers Insurance, shared insights into the company’s rapid deployment of AI technology, specifically in the critical first notice of loss (FNOL) process. With 28 years at Travelers and extensive experience in claims and technology, Eric oversees technology, analytics, and business delivery within claims, including AI initiatives. The FNOL process is vital as it sets the tone for the entire claims journey, helping customers navigate filing claims after incidents like accidents or storms, while also enabling operational efficiency by accurately capturing information upfront.
Travelers introduced an AI-powered claims assistant to handle high call volumes, especially during peak events like hurricanes, offering customers an option to interact with the AI for filing claims. This assistant guides users through the process, answers questions about coverage, deductibles, and fault, and can initiate related services such as scheduling repairs or rental cars. The AI system operates seamlessly with multiple agents working behind the scenes, providing a more adaptive and responsive experience compared to traditional chatbots, while always allowing customers to connect with human agents if preferred.
A key to Travelers’ success was adopting AI not just as a technology but as an operating layer, involving cross-functional teams from data engineers to legal experts early in the process. This collaborative approach, combined with rigorous change management and transparency, helped build trust and familiarity across the organization. They implemented a responsible AI framework aligned with their “claim three laws” to ensure accuracy, customer experience, and operational efficiency. Continuous iteration, testing, and feedback loops were integral, with daily updates and extensive evaluation to refine the AI’s performance.
Travelers also developed a “mission control” system to monitor AI performance in near real-time, supported by “LLM judges”—AI tools that assess tone, accuracy, and detect hallucinations or errors. This infrastructure enabled rapid scaling from an eight-state pilot to nationwide deployment within two months, achieving 80-90% customer adoption of the AI assistant. The partnership with OpenAI was crucial, characterized by close collaboration, responsiveness to feedback, and shared commitment to innovation, which accelerated development and ensured the AI met high standards of reliability and user satisfaction.
Looking ahead, Travelers plans to expand AI applications across other lines of business and throughout the entire claims lifecycle, with around 20 initiatives underway. The company emphasizes upskilling and reskilling employees to work alongside AI, viewing the technology as an enabler rather than a job threat. Eric expressed excitement about the ongoing partnership with OpenAI and the potential for AI to enhance claims processing by ensuring fair payments, excellent customer experiences, and operational efficiency, signaling a transformative future for the insurance industry.