JPMorgan, JPM, is reportedly consider reducing the ratio of junior bankers to senior managers from the current 6-1 to 4-1

According to CNBC, JPMorgan is evaluating a plan to change the ratio of junior analysts to senior bankers from 6:1 down to 4:1. But in that new structure, CNBC says half of the remaining junior staff would be placed in lower-cost countries like India or Argentina. So if you’re in New York or London, a two-thirds reduction may still end up being accurate.

The shift—unsurprisingly—is driven by AI. CNBC described how Derek Waldron, JPMorgan’s former chief analytics officer and ex-McKinsey partner, tested JPMorgan’s internal LLM by giving it this instruction:

“You are a technology banker at JPMorgan preparing for a meeting with Nvidia’s CEO and CFO. Create a five-page presentation with the latest news, earnings, and a peer comparison.”

Within 30 seconds, the model produced a polished PowerPoint deck. Waldron noted that this would normally require teams of analysts working late into the night.

And JPMorgan isn't alone. Many major banks are developing similar tools. Earlier this year, Goldman Sachs CEO David Solomon said their LLM could prepare 95% of an IPO prospectus in minutes.

Until now, banks haven’t been explicit about how many junior roles might disappear. JPMorgan’s internal discussions suggest a dramatic shift for major financial hubs. In this model, AI-enhanced junior teams in low-cost countries work around the clock to support senior bankers, lowering expenses and increasing profit margins. JPMorgan’s stock goes up, senior bankers focus on client relationships, and juniors in New York and London become ultra-efficient operators working alongside AI and offshore teammates.

That vision might excite consultants and AI evangelists, but it introduces obvious questions. Will clients be happy reading 50-page machine-generated pitchbooks? And where will future senior bankers come from if entry-level talent is outsourced or automated? Citi tried something similar years ago by offshoring analysts to Málaga—those analysts tried to get transferred to London until the unit was eventually phased out. As Solomon noted, when AI does 95% of the work, the remaining 5% handled by humans becomes critical—but only if those humans have learned the craft themselves.