Goldman partner warns AI could erode bankers' reasoning

Churchman's warning comes as Wall Street firms push AI deeper into trading and banking processes. Last year, CNBC reported that firms were exploring ways to use AI to reduce the ratio of junior bankers to senior employees — a step that could shrink the traditional pipeline for developing talent.

His concern is not about whether AI works, but about what happens when it handles routine work that historically trained young professionals. He points to junior traders learning by fielding client pricing requests under supervision, noting 'we can absolutely automate that, but then do we get the senior traders that fully understand?' This reflects a broader tension between short-term efficiency and long-term institutional knowledge.

Even Goldman, one of the world's top investment banks, has not 'figured out' how to manage the transition it has begun. The firm's experience with Marquee suggests the technical hurdles — particularly ensuring AI answers are factual and auditable — remain significant. In high finance, where tolerance for errors is low, Churchman's 'sounding thorough' admission underscores that AI systems still need human oversight in high-uncertainty decisions.