2026 Agenda

Compliance activities in financial services and insurance remain dominated by manual checks and siloed rule engines, despite increasing transaction volumes and regulatory complexity. This keynote explores a design‑science research framework that combines deterministic regulatory rules, machine‑learning risk detection and explainable AI into an adaptive compliance enforcement system. The proposed architecture spans six layers, from data ingestion through risk scoring, explainability and a validate‑before‑delivery enforcement engine to supervisory dashboards. We simulate the framework on a synthetic dataset mirroring 50,000 real‑world transactions and compare it against a manual baseline. Results show marked improvements – detection rate rises from roughly 70 percent to 90 percent, false positive rates halve, latency is cut by half and manual review hours drop from eight to three per day. The talk highlights how combining rules and AI delivers transparent, trustworthy and scalable compliance, discusses limitations such as bias and the need for investment, and outlines directions for piloting and adaptive learning.