Goldman Sachs partner Chris Churchman warned on August 24, 2026, that the rapid deployment of artificial intelligence across Wall Street risks triggering cognitive atrophy in the next generation of financiers, according to CNBC. Speaking on the firm’s Exchanges podcast, Churchman stated that delegating core analytical reasoning to automated models threatens to erode the fundamental problem-solving skills junior bankers need to build long-term careers.
The Risk of Cognitive Atrophy in Banking
According to Chris Churchman, who leads Goldman Sachs‘ digital platform for institutional clients called Marquee, the financial industry faces a profound structural paradox as it adopts automated systems. While AI tools boost immediate efficiency and profitability, they simultaneously eliminate the routine grunt work that traditionally taught junior employees how to think critically, structure arguments, and reason from first principles. Churchman noted that just as modern inventions diminished human navigation and memorization capabilities, algorithms threaten to replace human analytical rigor if bankers stop exercising those mental muscles themselves. The co-chair of the firm’s Global Banking and Markets AI working group emphasized that reasoning remains essential for structuring complex arguments, warning that current professionals are increasingly delegating that core function to software.
Apprentice Culture and Workforce Restructuring
The shift toward automation also threatens Wall Street’s traditional apprenticeship model, where much critical knowledge is passed down informally through hands-on experience, according to CNBC. Churchman pointed out that junior traders historically learn their craft by fielding client pricing requests under the direct supervision of experienced risk takers. Although banks can automate those pricing tasks entirely, Churchman questioned whether automated workflows will produce future senior traders who truly understand market mechanics. Goldman Sachs has not yet solved how to manage this transition while preserving the tacit knowledge embedded in its existing workforce, according to Churchman.
Technical Hurdles in Financial AI Platforms
Beyond workforce impacts, developing enterprise-grade financial AI presents severe technical hurdles regarding factual accuracy and auditability, according to CNBC. Churchman, who previously ran currency trading at UBS before joining Goldman Sachs in 2021, shared implementation lessons from Marquee, the bank’s digital platform providing institutional clients with research, risk analytics, and trade execution. While consumer-facing AI tools often rely on disclaimer warnings about potential mistakes, high-stakes financial environments demand near-zero tolerance for errors. During the development of Goldman’s internal AI platform—which remains restricted to bank employees—testing revealed stark limitations in model reliability. When developers challenged the system rigorously, the software made an explicit admission regarding its operational limits, with Churchman quoting the model as stating, Look, in the end, I’m better at sounding thorough than being thorough.
Keep reading