The math doesn’t work the way the banks seem to think it does.
Goldman Sachs, JPMorgan Chase, Citigroup, and Barclays are reported to be cutting junior analyst classes as AI-driven tools take over the workflow functions that entry-level analysts have traditionally owned: pitch books, financial models, comparable company analysis. According to Straits Times reporting, reductions at some institutions have reportedly reached as much as two-thirds of prior cohort sizes. That’s the maximum claim in available reporting, not a confirmed average across all named institutions. Standard Chartered has reportedly announced plans to eliminate approximately 7,800 roles by 2030 as part of a broader AI-driven restructuring, per the same reporting. That figure is a single-source claim pending additional corroboration.
The executive framing is consistent across institutions. Goldman Sachs President John Waldron has reportedly described the firm’s existing workflows as resembling a “human assembly line” suited for digitization, according to Business Insider reporting. Citigroup CEO Jane Fraser and JPMorgan CEO Jamie Dimon have both made public statements indicating AI-driven role elimination is a planned development. These are attributed statements, not verified against primary transcripts in available sources, they’re consistent with public positioning from both institutions, but they carry “reportedly” framing here.
Three CEOs. Same message. That’s a pattern, not a coincidence.
The catch is what those junior analyst classes represent beyond entry-level headcount. In investment banking, the analyst cohort is traditionally the primary pipeline for quantitative roles, data science positions, and eventually the internal AI development teams that banks increasingly depend on. The logic of the current restructuring assumes that AI replaces the analytical work, but it doesn’t automatically generate the engineers who build, maintain, and customize the AI doing the replacing. Those engineers have to come from somewhere.
This story is an ongoing structural trend, not a single-date event. The Straits Times reporting reflects a pattern that has been developing across multiple reporting cycles and isn’t tied to a single announcement. The Challenger data brief published June 6 documented that AI-attributed job cuts in 2026 have already exceeded the full-year 2025 total across all sectors. The hub’s stakeholder analysis on sector-wide AI cutting patterns provides the broader context for what’s happening in finance specifically.
What to watch
the WARN Act dimension. The federal AI bill published June 7 includes a provision that would create new employer obligations for AI-driven layoffs. Banks reducing cohorts rather than conducting mass layoffs currently avoid WARN Act thresholds, but regulatory attention to the practice is building. Whether reduced analyst classes count as AI-driven displacement under emerging definitions is an open compliance question. The regulation pillar covered this directly on June 7.
The real story is the talent pipeline paradox. Banks are automating the entry-level roles that historically produced their internal AI talent. If the automation works as intended, they’ll need more AI capability in two to three years, not less. Who builds it if the cohort that would have built it wasn’t hired?