NEW YORK — The Nasdaq Stock Market index is using AI to streamline its efforts to investigate suspicious trading and anti-money laundering.
The New York-based exchange is using gen AI to identify behaviors like pump-and-dump or spoofing schemes in trading, Tony Sio, head of regulatory strategy and innovation at Nasdaq, said at this week’s AWS Financial Services Symposium in New York City.
The gen AI model, built with Amazon Bedrock and SageMaker, helps analysts investigate suspicious trades, reduces false positives, and can speed the investigation process by nearly 33%, Sio said.
“We’re using gen AI to essentially review corporate documentation provided by firms and what is in the news,” Sio said. “Gen AI can intelligently [provide] the investigative evidence that the analyst uses to determine if something is fraud or not fraud.”
The gen AI-aided investigation model helps analyze investigations by condensing information from specific sources, including:
- Historical price and volume analysis;
- Company news and summaries;
- Sector peer news;
- Corporate filings; and
- Company shareholder news.
The model is designed to flag suspicious trades and provide supporting documents in an investigation rather than making decisions, Sio said.
Nasdaq worked with AWS for 15 months to develop the model, which was rolled out in the U.S. and will be rolled out in the EU later this year, Sio said.
NYSE deploys gen AI for market surveillance
Meanwhile, the New York Stock Exchange is also deploying gen AI for market surveillance with a news summarization tool built using Amazon Bedrock to deliver deeper, more relevant insights to analysts.
The tool helps reduce the manual labor required to investigate the fluctuations of stock prices and potential foul play, and has attained an accuracy rate of 98% in its investigations, Anand Pradhan, senior director of head of regulatory and network management system technology at New York Stock Exchange, said at the event.
NYSE executes “650 billion transactions on a daily basis,” and gen AI can help flag and investigate suspicious trades, making market surveillance more efficient, Pradhan said.
