SCOR Launches GenAI for Underwriting and Claims

SCOR's AI Assistant platform is engineered to ingest and reason with unstructured medical documents from multiple languages and markets. It creates a standardized "digital twin" of the applicant's essential information for underwriters, which is then integrated into its automated underwriting system, Velogica, and its new claims system, VClaims. The system is designed with multiple validation layers to ensure precision and keeps human experts in control of the final decision-making process. This move reflects a broader industry trend of leveraging GenAI to analyze unstructured data like medical reports, broker emails, and loss runs, which have historically been difficult for traditional models to process. The goal is to augment, not just automate, by allowing AI to handle initial data extraction and risk scoring, freeing up human underwriters to focus on complex risk assessment. For claims, this translates to faster damage assessment from photos and more sophisticated fraud detection by analyzing patterns across large datasets. Architecturally, such systems often employ a multi-agent approach where specialized AI agents handle distinct tasks like data extraction, risk classification, or compliance checks. These agents are managed by an LLM orchestration framework, which sequences tasks, manages communication between agents, and integrates human-in-the-loop feedback. This modular, often graph-based, design is crucial for building auditable and reliable workflows in regulated industries like insurance. Connecting these AI capabilities to legacy core systems is a primary challenge, with 85% of insurers citing it as a major barrier. An API-first strategy is becoming standard, using RESTful APIs to encapsulate legacy platforms and expose their data to modern, cloud-native services without a complete system overhaul. This enables real-time data exchange and allows insurers to integrate third-party data sources and digital tools more effectively. For insurtech startups, this technological shift creates new opportunities, but the funding landscape has changed. After several years of contraction, venture investment in insurtech is stabilizing, with a projected modest increase for 2025 over 2024. However, investors are now more selective, favoring companies with proven technology and clear paths to profitability over hyper-growth. Late-stage startups now attract 60% of insurer investments, a significant increase from 25% in 2023. The emphasis on explainability is paramount for adoption and regulatory compliance. Systems must be able to surface the "why" behind their risk scores and decisions in natural language. This involves robust semantic record-keeping of AI inputs, outputs, and intermediate steps to ensure auditability. For technical founders, this means building not just powerful AI, but also trustworthy systems with strong governance and role-based access controls to prevent data leakage.