Public debate remains fixated on predictions of AI-driven mass white-collar unemployment, a narrative recently reinforced by Anthropic CEO Dario Amodei’s repeated warnings of imminent labor displacement.

However, Anthropic’s own Chief Economist Peter McCrory published findings based on 18 months of Bureau of Labor Statistics data and internal Claude usage logs that challenge those alarmist forecasts. The data shows that the U.S. unemployment rate among highly exposed knowledge workers has not meaningfully diverged from unexposed roles, largely because AI functions as an iterative, collaborative tool rather than a full job substitute due to its jagged capability profile.

While aggregate job loss has not materialized, early warning signs are appearing in early-career hiring, where job postings for entry-level positions in high-exposure technical fields have contracted noticeably.

In federal technology and defense contracting, this macro dynamic is playing out in reverse, as agencies expand specialized civil service headcount while pushing contractors to integrate AI directly into complex supply chain and logistics enclaves.

What's Inside

Let's get into it.