SCOR Deploys GenAI for Underwriting
Global reinsurer SCOR launched a GenAI-powered solution for its life and health clients to accelerate medical underwriting and claims decisions. The system uses Optical Character Recognition (OCR) modules to extract data from medical records and supports Verisk’s EHR Triage Tool in the U.S. to leverage electronic health records for faster risk assessment.
- SCOR's proprietary system is named AI Assistant; it was piloted internally in 2024 and is now used daily by over 150 of the company's underwriting and claims experts. - The tool currently processes around one million pages each month with approximately 90% accuracy across key data fields, and has already achieved time savings of 30% for medical underwriting in Life & Health. - The collaboration with Verisk on its EHR Triage Engine is a long-standing partnership that began in August 2020. The engine is capable of evaluating 95,000 medical impairments in five minutes or less, which can help qualify as many as 85% of applicants with minimal to no human review. - The GenAI tool is part of a larger strategic platform called SCOR Digital Solutions, which also houses its automated underwriting suite (Velogica), a digital claims solution (VClaims), and a predictive modeling solution (BeeTech). - This initiative is a core component of the reinsurer's "Forward 2026" strategic plan, which is focused on digital transformation to make the company future-ready. - According to SCOR's Chief Data & Analytics Officer, Sofia Kyriakopoulou, the approach is to keep a "human in the loop" to validate the information extracted by the AI and augment their employees' expertise. - The GenAI solution will be integrated with SOLEM, SCOR's existing underwriting manual, to pre-select decision proposals in rating grids based on the information captured from medical evidence. - Beyond its own operations, SCOR has also joined a global insurance facility to provide coverage for AI and data center infrastructure, acting as the lead insurer for the environmental impairment liability component.