Goldman Sachs’ Chris Churchman warns AI could dull junior bankers’ reasoning skills

Goldman Sachs partner Chris Churchman, who leads the firm’s Marquee platform, is warning that the rush to use artificial intelligence at investment banks could come with a skills cost. In comments carried by Yahoo Finance, Churchman cautioned that if junior bankers outsource too much thinking to AI tools, they may lose the habit of independent reasoning and judgment, a problem he described as “cognitive atrophy.”

Churchman’s remark centers on how AI is used in day-to-day work in banking teams. AI systems can draft, summarize, and generate working materials quickly. But the executive’s concern is less about whether AI can produce outputs, and more about whether it reduces the practice of building expertise, stress-testing ideas, and turning incomplete information into decisions.

The warning comes as major Wall Street firms experiment with AI in areas ranging from research and document processing to coding and drafting. For investment banks, the impact is particularly sensitive at the junior level because training and mentorship are tied to developing instincts and analytical muscle over time. Churchman’s framing implies that the benefits of AI efficiency need to be balanced against the long-term development of talent.

Goldman’s Marquee platform, which Churchman leads, is the type of internal technology effort that typically brings AI tools closer to bankers’ workflows. While the post described his role in leading the platform, it did not provide additional technical specifics in the material reviewed here, such as which models are used, how they are governed, or which tasks are restricted.

Churchman’s message also reflects a broader debate inside financial services about AI use policies. Firms generally emphasize controls, including review processes and approvals for outputs used externally. Yet the concern he raised suggests that governance alone may not solve the training issue, because even compliant tools can still change how analysts learn if they become default substitutes for problem-solving.

Goldman Sachs shares are traded on the New York Stock Exchange under the ticker GS. The firm’s business spans investment banking, trading, and asset management, and AI adoption is likely to be uneven across those lines, depending on data availability, regulatory scrutiny, and operational risk. Churchman’s comments, though focused on skills, implicitly point to how AI rollout affects culture as much as productivity.

Still, the disclosure in the published comments is high-level. The post did not quantify how common AI-assisted workflows are within Goldman, did not describe any internal program or specific policy changes tied to the warning, and did not clarify whether Marquee is designed to encourage “human-in-the-loop” reasoning or whether it could be tuned to require more analyst work before outputs are produced.

Going forward, investors and observers will likely watch whether Goldman elaborates on concrete training guidance, AI usage guardrails for junior staff, or measurable education outcomes tied to AI tools. The key question will be whether the firm pairs faster generation with disciplined reasoning habits, rather than simply accelerating who can produce a first draft.