Goldman Sachs and Deutsche Bank test agentic AI for trade surveillance

Goldman Sachs and Deutsche Bank are testing advanced "agentic" AI systems for trade surveillance that go beyond traditional rule-based monitoring by analyzing trading patterns in real time and reasoning through complex data to identify potentially suspicious conduct. Unlike conventional automated systems that rely on static alerts triggered by predefined thresholds, these adaptive AI agents evaluate multiple signals, compare current activity with historical data, and detect subtle anomalies indicative of misconduct, thereby enhancing oversight in increasingly complex and high-volume markets. This development aims to improve the accuracy of trade monitoring, reduce false positives, and better identify nuanced forms of market manipulation without replacing human compliance teams.