A screen displays the Goldman Sachs company logo on the floor of the New York Stock Exchange (NYSE) in New York City, U.S., on May 7, 2025. Brendan McDermid | A Reuters partner, Goldman Sachs, who leads one of the bank’s flagship artificial intelligence projects, warned that the spread of AI across Wall Street risks hampering the thinking capabilities of the next generation of financiers. “There’s a huge danger here that in the age of AI, we outsource our reasoning to these models, and we have cognitive atrophy that prevents us from reasoning from first principles ourselves,” said Chris Churchman, who runs Goldman’s digital platform for institutional clients called Marquee. The comments came during the latest episode of the report. “Exchanges” of the firm, according to a transcript provided exclusively to CNBC. Just as people lost navigation and memorization skills with modern inventions, bankers risk losing analytical capabilities if algorithms handle all the heavy lifting, Churchman said. “Reasoning is still important,” he said. “You still need to reason about [problems] and structure it into an argument, and now we’re delegating the reasoning.” Wall Street’s push to integrate AI into all of its business and banking processes could be something of a devil’s pact: It will make the industry more profitable today and potentially erode the talent it needs for tomorrow. With AI taking on more of the routine work that has traditionally taught young bankers and traders how to think and make decisions, companies risk sacrificing the culture that turns young employees into seasoned Wall Street talent. Even could reduce the need for junior bankers first, CNBC reported last year that Wall Street firms were examining ways to use AI to reduce the ratio of junior bankers to senior employees. Banks need to find a balance between using AI and preserving Wall Street’s learning culture, said Churchman, who headed currency trading at UBS before joining Goldman in 2021. “You learn by doing, and a lot of knowledge is tacit, it was never written down.” tacit and intuitive that some of our best people have today. [and] For example, junior traders learn by answering clients’ price requests under the supervision of experienced risk takers, Churchman said. “We can absolutely automate that,” he said, “but will we get senior traders to fully understand?” Systems should be designed so that employees still call the shots in high-risk, high-uncertainty decisions rather than becoming passive operators, Churchman said. He hasn’t yet “figured out” how he will manage the transition the company has begun, said Churchman, who is also co-chair of the firm’s Global Banking and Markets AI working group. No errors? Also in the podcast interview, Churchman shared lessons from implementing AI at Marquee, which is used by hedge funds and other institutional clients to access Goldman’s market data, research, risk analysis and trade execution services. The Marquee AI platform is only available to Goldman employees for now, he said. The most difficult challenge, from a technical point of view, is ensuring that AI responses are 100% objective and can be audited, he said. While consumer AI chatbots warn users of potential errors, in high finance, error tolerance is low. Churchman said that in developing the company’s AI platform for clients, the software made a surprising admission: “When we challenged him a lot, at least he was honest,” Churchman said. sound thorough than being thorough.'”Choose CNBC as your preferred source on Google and never miss a moment from the most trusted name in business news.
