Anthropic says junior technical hires fell to 9% from 27%

Anthropic has sharply reduced the share of junior engineers and researchers it hires while its use of AI agents has surged, according to a report by company researchers published Friday. Junior staff made up 14% of the company’s technical departments, down from 21% in early 2025. Junior roles fell to 9% of technical hires, from 27%.

The report covers one company, and the company sells the AI tools it studies. But Anthropic says its heavy use of agents could make it “a leading indicator” of how AI shifts hiring across the economy.

The numbers: The report draws on internal data on employment, AI use, output, surveys, calendars and hiring records, plus 17 employee interviews. Its authors are Nathan Wilmers, Regina Catipon, Maxim Massenkoff and Peter McCrory.

Postings for junior jobs have fallen “particularly sharply” since December 2025, the report says. By April, engineering and research teams had deemphasized hiring for lower-level technical roles. About a third of machine learning managers and a fifth of software engineering managers said they were hiring less because of AI.

Technical hiring still grew in absolute terms. But sales, finance and recruiting grew faster, and technical departments’ share of staff fell to 56% from 61% over the past year.

Agent work hours passed human work hours in March 2026, the report says. By August, agents directed by individual employees logged six hours for every one their human users worked. The report counts agent runtime, excluding waits longer than 30 minutes, and assumes 40-hour human workweeks.

Code commits per engineer ran at five times their pace of six months earlier. The researchers tracked 192 engineers who adopted coding agents and 74 who did not. They found no difference in how often agent-written code was later removed or reverted.

What it means: The report links the hiring shift to AI but does not prove it. The company grew fast in the same period, and the authors note that adoption was not random: engineers ramping up big projects may have been the first to adopt agents.

Human attention has become the bottleneck, the report says. Agents in interactive sessions spent about 80% of their time waiting for human input, the report says. Hiring managers now put more weight on autonomy and judgment than on technical expertise.

Employees expect AI to be able to do 75% of their work by 2030, compared with 28% for other Americans, according to company survey data cited in the report.

What’s next: Anthropic says it has set up an early warning system to track jobs at risk of disruption and is prioritizing internal moves. Employees with at least a year at the company were 50% more likely to switch departments each month in 2026 than in 2025. The report says those steps “raise difficult questions” about what such commitments should be.

Sources: Anthropic, “How employment is changing at Anthropic”