How AI Agents Are Replacing Junior Knowledge Workers

The displacement of junior knowledge workers by AI agents is no longer a hypothetical scenario appearing in think-tank reports. It is a trend visible in hiring data, law firm staffing memos, and the quarterly earnings calls of professional services companies that have quietly reduced graduate intake while maintaining or growing revenue.

At a mid-size US law firm, paralegals who previously spent 60 percent of their time on document review and contract summarisation have seen that work handed to an internal AI system. The firm's head of technology describes it as a force multiplier for senior lawyers — the same partners can now handle roughly 40 percent more client engagements. The junior headcount, however, has not grown in three years despite record revenue.

In financial services, the pattern repeats. Equity research desks that once employed large teams of analysts producing first-draft earnings models now route that work through LLM-based agents that can process a 200-page annual report, extract key figures, and produce a formatted model in under four minutes. The remaining human analysts are expected to operate at a higher level of synthesis and client interaction from day one.

The question of what this means for career pipelines — where junior roles have traditionally served as training grounds for senior expertise — is one that neither firms nor policymakers have seriously answered. The productivity gains are real and measurable. The structural consequences for talent development are equally real, and considerably less examined.