Home / Business and Economy / Banks Harness AI for Trading Surveillance
Banks Harness AI for Trading Surveillance
25 Feb
Summary
- AI is being integrated to monitor client-facing staff communications.
- Deutsche Bank uses AI to reduce false positives by over 25%.
- Banks aim to improve efficiency and cut compliance costs with AI.
Financial institutions are actively integrating artificial intelligence to boost efficiency and reduce costs, with a significant focus on enhancing trading surveillance. Deutsche Bank and Goldman Sachs are pioneering the use of agentic AI to monitor client-facing staff and detect potential misconduct.
Deutsche Bank is collaborating with Google Cloud to develop large language models capable of identifying anomalies in orders and trades. This initiative aims to flag potential market abuse for human review, with plans to monitor staff communications later this year. The bank has already retired legacy systems, leading to a reduction in false positives by over 25% and shutting down 200 internal servers.
Nomura Holdings Inc. is also exploring collaborations to train AI surveillance models together, potentially reducing false positives by 30% to 40% and saving millions in compliance costs. While AI is proving effective, banks are proceeding cautiously, ensuring human oversight remains integral to the process. This phased approach aims to mitigate risks associated with new vulnerabilities and sensitive data exposure.
Additionally, firms like ThetaRay are assisting banks such as Banco Santander SA in improving anti-money laundering controls using agentic AI. Despite advancements, the ultimate decision-making authority remains with human compliance officers, ensuring a balance between technological efficiency and human judgment.
