By Mari Hansen The analyst bullpen, the small row of desks where new hires used to bump into each other for screen time on the trading floor of a major investment bank in lower Manhattan, has quietly shrunk. Not in a big way. Not through a press release. There are just fewer chairs, fewer headsets, and fewer faces from Wharton squinting at pitch decks at two in the morning. This summer’s interns were given tasks that he used to assign to associates, a vice president who has been there for eleven years told me with a hint of amusement. This thing is shaped like that. It’s not coming with a siren.
The next wave of AI displacement is quietly taking hold in the finance sector, and banks are far ahead of their public messaging. In the past eighteen months, JPMorgan, Morgan Stanley, Goldman Sachs, and Citi have all implemented internal large-language-model platforms that were trained on client data, proprietary research archives, and filings. These are described to bankers as “productivity tools.” Technically, that is correct. Every automation wave in history has been introduced in this manner as well. In the financial services industry, the productivity gain typically results in smaller hiring classes rather than fewer hours for current employees.
| Category | Details |
|---|---|
| Primary Trend | AI-driven displacement of entry- and mid-level white-collar finance roles |
| Key Prediction Source | Dario Amodei, CEO, Anthropic |
| Amodei’s Claim (2025) | AI could eliminate up to 50% of entry-level white-collar jobs |
| Major Research Source | University of Pennsylvania / OpenAI study published in Science (2023) |
| Workers With at Least 10% Task Exposure | Roughly 80% of U.S. workers |
| Workers With 50%+ Exposure | Approximately 19% |
| Banking-Sector Study Reference | Goldman Sachs — up to 300 million global jobs task-automatable |
| Most Exposed Finance Functions | Junior analysts, accountants, risk processors, compliance analysts |
| Key Academic Tracking Body | MIT Work of the Future; Stanford HAI AI Index |
| Least Exposed Finance Roles | Senior bankers, client advisors, strategy executives |
| Regulatory Watchdog Monitoring Impact | U.S. Federal Reserve labor market reports |
| Current Employee Resistance | Roughly 80% of enterprise workers avoiding or rejecting AI tools (Fortune, April 2026) |
Academic research is catching up to Wall Street’s existing understanding. About 19% of American workers have at least half of their tasks exposed to current AI capabilities, according to a 2023 study published in Science. Finance analysts, accountants, compliance reviewers, and paralegals routinely rank near the top. A quarter of all work worldwide could be automated, according to Goldman Sachs’ own widely cited report. It’s a little ironic that the same banks that are discreetly restructuring around displacement are also the ones publishing research on it.
The CEO of Anthropic, Dario Amodei, made headlines last year when he said that 50% of entry-level white-collar jobs might vanish in a few years. The figure is contoversial, and his detractors are correct to point out that alerting the public about your own product is a marketing tactic. However, the number doesn’t seem ridiculous to those in charge of bank operations.
It sounds similar to the current state of prettier math. Diligence memos used to take weeks for junior associates to complete; today, AI tools draft them in the afternoon, and senior bankers edit them. To feed that pipeline, fewer juniors are required. It’s still unclear if the new baseline is just thinner or if the work will resume when AI tools reach a plateau.
The atmosphere within the companies is peculiar. No obvious panic is present. Roughly 80% of enterprise workers are either avoiding or reluctantly using AI tools, which is a quiet rebellion in and of itself, according to a recent Fortune article. Analysts remain focused. The way managing directors discuss “leverage ratios” has changed. In the meantime, hiring freezes are being referred to by HR departments as “reshaping”—a term that requires a lot of work.
This is a cultural component that the industry has yet to fully comprehend. For aspirational graduates, finance has always been the career escalator—long hours, demanding apprenticeships, but a dependable route from cubicle to corner office. The entire structure becomes weaker if the first three rungs of that ladder are automated away. It is impossible to develop senior bankers without cultivating juniors. As this develops, there’s a sense that the banks are aware that the outdated model is failing, but no one wants to be the first to publicly declare this. A headcount committee is running the numbers for next year somewhere in Midtown. The silent portion won’t remain silent for very long.