The Anthropic logo. Image: Anthropic
Anthropic builds the AI that is supposed to change white-collar work, so it has now looked at what that AI has done to its own staff. Junior employees made up 21% of its technical headcount in early 2025 and 14% now, and the share of technical hires who are junior fell from 27% to 9%, according to “How employment is changing at Anthropic”, a report published on October 16.
The report, by economists Nathan Wilmers, Regina Catipon, Maxim Massenkoff and Peter McCrory of the Anthropic Institute, draws on internal records of headcount, AI use, output, surveys, calendars and hiring decisions, plus interviews with 17 employees. Its authors pitch Anthropic as a possible “leading indicator” for the rest of the economy.
Fewer juniors, from January
Talk of falling demand for junior technical staff began in Anthropic’s core technical teams in January 2026, the report says, as the teams leaning hardest on Claude found that writing code and finding bugs could be handed to it. By April, engineering and research teams had “deemphasized” hiring for lower-band technical roles. In a survey that month, around a third of machine-learning managers and a fifth of software engineering managers said they were hiring less because of AI.
Postings for junior roles have fallen especially sharply since December 2025. Technical departments also shrank as a share of the company, from 61% to 56% over the year, though technical hiring is still growing in absolute terms: sales, finance and recruiting simply grew faster. Anthropic has tripled in size in the last year.
Administrative work barely grew at all. One executive assistant told the researchers: “I think we’ve already moved past the point of Claude augmenting away the really entry-level admin type role.”
Agents now outwork people 6 to 1
In March 2026 the combined hours worked by Anthropic’s autonomous and interactive AI agents overtook human hours for the first time. Counting only the agents that individual employees direct themselves, by August they were working six times as many hours as the people running them.
Output rose with it. Code commits per engineer are five times what they were six months ago. Following 192 engineers who adopted coding agents against 74 who did not, the authors found the code they added rose sharply after adoption, with no rise in how often that code was later deleted or reverted. Lawyers processed more contracts, and salespeople closed more deals. Non-technical staff with no code commits of their own went from 2 coding agent sessions a month in August 2025 to 90 in May 2026.
“An army of Claudes, with every one asking for my attention”
That is how one engineer described the job now. The report finds Claude sessions spend around 80% of their time idle, waiting on a human. Even for non-technical staff, simply reading everything their agents produce would take more than half the working week. “Meta-work”, such as writing specs for agents, setting one agent to check another and watching pipelines, rose from 10% of the average employee’s Claude sessions in autumn 2025 to 22% in spring 2026.
What managers hire for has shifted too. Interviewers increasingly comment on candidates’ autonomy and judgement, and less on domain expertise. Mentions of a candidate’s track record with AI tools tripled for technical roles between 2024 and 2026, and quadrupled for non-technical ones.
The figures sit alongside Anthropic’s “AI expectations gap” survey from the same week. The new report says Anthropic staff expect AI to be able to do 75% of their work by 2030, against 28% for other Americans.
What Anthropic says it is doing
The company has set up an early warning system that reads role-level demand and productivity against AI use to flag jobs at risk, and it is prioritising moving people internally: staff with at least a year at the company were 50% more likely to switch departments each month in 2026 than in 2025. The authors also point to widespread employee share ownership.
They are frank about the limit. None of these steps “directly address the slowdown in junior hires documented in this report”, they write, adding that “employment protections for incumbents will not benefit those not hired to begin with”. They call for government policy on retraining and income support. Earlier this autumn, Apple drew up plans to swap 5,000 support staff for AI agents before putting them on hold.
Why it matters
This is the first time a frontier AI lab has published its own data showing entry-level hiring shrinking as its AI takes on the work. If Anthropic really is a leading indicator, the first rung of the technical career ladder is the one to watch. The figures are Anthropic’s own and have not been independently checked.
Sources: Anthropic, “How employment is changing at Anthropic” (Wilmers, Catipon, Massenkoff, McCrory, October 16, 2026).
Latest Anthropic news
