AI Driven Fraud Probability Model for Open Claims Qorus-NTT DATA Innovation in Insurance Awards 2026

This project introduces an AI‑driven fraud probability model that analyzes all open Home and Commercial claims and assigns each of them a real‑time fraud‑risk score. The objective is to identify suspicious cases earlier, allow investigators to focus on the highest‑risk claims, and significantly reduce operational leakage across the claims journey. The model transforms traditional rule‑based fraud detection into a proactive, predictive, and data‑driven process fully integrated into claims operations.

Additionally, the project leverages the machine‑learning algorithm already developed by BU Spain, but fully adapts and extends it to the specific behavioural patterns, operational workflows and fraud dynamics of the Joint Venture. While the original algorithm provided a strong technical foundation, the final model represents a fully customized solution built around our own claim portfolio and Investigation Unit insights. This combination of a solid corporate foundation with local, tailored innovation is one of the key differentiating factors of the project.