AI AML Agent (Anti-Money Laundering) Qorus-NTT DATA Innovation in Insurance Awards 2026

The innovation consists of the development and implementation of a generative AI agent whose primary objective is to conduct due diligence in a standardized, secure, automated, and contextualized manner within Anti-Money Laundering (AML) processes.

The project originated from a critical need to scale operations in response to the exponential growth of the customer base in 2024, which generated a volume of risk events and pending due diligence cases that exceeded manual processing capacity. Unlike conventional automation solutions available in the market, this agent was designed to interpret information contextually and semantically, going beyond simple keyword searches to understand the actual meaning of news reports, restrictive lists, and politically exposed persons (PEP) profiles.

The solution operates as a multi-agent system capable of processing complex documents, searching for matches in regulatory databases, calculating automated risk scores, and generating auditable reports with forensic traceability. In doing so, the agent began to support the due diligence analysis process in a structured manner, contributing to a 63% productivity gain in operations. In addition, tests demonstrated a high level of reliability, with 90% convergence between the analyses performed by the agent and those conducted by employees. In the remaining cases, the agent demonstrated a more conservative risk assessment approach.

The development involved cross-functional collaboration between the Innovation, AI Engineering, Architecture and Information Security, Infrastructure, and Compliance teams, ensuring a robust, secure, scalable, and auditable architecture. As a result, the project demonstrates significant productivity gains and increased technical rigor in several scenarios, consolidating artificial intelligence as a key pillar that combines regulatory compliance with operational agility.