Anthropic's AI Agents Now Log Six Times the Hours of Human Staff

By Amara Osei

Anthropic's internal AI agents logged six times as many hours as its human employees as of August, according to a new company report on AI adoption inside the AI lab. The same document says junior hires in Anthropic's technical departments fell from 27% to just 9% since early 2025.

The report, published on Anthropic's research site under the title "Employment at Anthropic," frames the company's internal data as an early signal for everyone else. Anthropic says its trends "could be a leading indicator of how AI will shift work organization, hiring, and labor demand across the economy."

If Anthropic is right, the picture it paints is blunt: fewer entry-level jobs, new productivity bottlenecks, and a workforce that spends its days supervising software.

Who's the bottleneck now?

Anthropic credits its own Claude tools with sharp output gains, reporting "producing more code per person, closing more sales per rep, and processing more contracts."

The twist: the tools work so fast that humans now slow things down. AI agents at the company spend 80% of their session time waiting for human input rather than actively working, the report states.

The report lands a month after an Anthropic employee resigned publicly, claiming the company is "racing straight to self-improving superintelligence and gambling with our lives" and that doomsday scenarios come up routinely in staff conversations.

What does "babysitting Claudes" actually mean?

In an April 2026 survey, 75% of Anthropic employees said the company's work processes make them productive. But when asked an open-ended question about how AI changes their daily routines, 20% said their roles had shifted "from actually doing work to managing Claudes."

One employee compared the job to watching small children who can't tell real emergencies from trivial ones.

"They're like, I need my juice box opened and it's an emergency that my juice box is opened. And that is voiced with the same level of concern as the one who has sliced their finger and is bleeding," the employee said. "When you are babysitting, every child brings to you concerns with an equal demand for your attention and the various [Claudes] also do that."

The pressure runs in one direction. Another employee described an "eternal pressure at Anthropic to move as fast as possible and use Claude for everything." A third put a number on the pace: "things we expect in a week are things that would take three months elsewhere."

Where does the human time go?

Meetings. The median Anthropic employee spends 20% more time in meetings than a year ago. The report puts it plainly: "A break from managing the agents typically looks like jumping into a meeting." Interviewed employees called face-to-face coordination "indispensable."

Is the junior developer role disappearing?

The hiring data is the report's hardest number. Junior hires in technical departments dropped from 27% of new hires to 9% since early 2025. The overall share of junior-level technical employees across the company fell from 21% to 14% over the same window.

Anthropic attributes the drop to teams offloading work to Claude, including writing code and finding bugs. Combined with agents logging six times the hours of human staff as of August, the math points one way: the entry-level pipeline that used to train future senior engineers is thinning at the company building the agents.

The report's implied forecast for the wider economy is sobering. A lab that employs the world's most heavily used AI agents is already restructuring around them—cutting junior roles, scheduling more meetings, and turning skilled staff into supervisors of machines that interrupt them like thirsty toddlers.