Deutsche Bank and Goldman Sachs are exploring the use of agentic artificial intelligence to enhance trade surveillance and track potential misconduct, indicating that financial institutions are increasingly integrating such technologies into their daily operations.
Bernd Leukert, Deutsche Bank's Chief of Technology, Data, and Innovation, stated that the bank is collaborating with Google Cloud, a unit of Alphabet Inc., to develop a large language model designed to detect anomalies in orders, trades, and market fluctuations.
Agentic artificial intelligence generally refers to AI systems capable of autonomous planning and action, differing from chatbots that merely provide information. Deutsche Bank aims for this tool to flag potential market abuse and escalate cases to human compliance staff once it becomes operational.
Additionally, Deutsche Bank plans to use large language models to monitor communications among traders, sales personnel, and other client-facing employees. This deployment is expected to launch later this year.
Meanwhile, according to informed sources, Goldman Sachs is also researching the use of agentic artificial intelligence to analyze trades and identify suspicious signals or unusual market movements. The sources requested anonymity due to the discussion of undisclosed information. A representative for Goldman Sachs declined to comment.
At Nomura Holdings, management is in talks with another global bank to jointly train an AI monitoring model, as revealed by Tahir Zafar, the firm's International Head of AI Strategy.
Many banks are evaluating ways to integrate artificial intelligence to reduce costs and improve efficiency. Currently, most trade monitoring still relies on rule-based algorithms that identify issues through predefined procedures.
Comments