What’s it really like to work at Anthropic? The AI company made headlines last month when an employee resigned, claiming the company is “racing straight to self-improving superintelligence and gambling with our lives” and that conversations about an AI-powered doomsday are par for the course among its staff.
Now, Anthropic is pulling back the curtain on its own workplace with a new report on the impact of AI adoption at the company. The company says its internal trends “could be a leading indicator of how AI will shift work organization, hiring, and labor demand across the economy.” If it’s right, that means a world with fewer entry-level jobs, new roadblocks to productivity, and a workforce full of babysitters for AI agents.
The babysitters club
Anthropic reports that AI is responsible for massive booms in productivity, saying its internal use of Claude tools has led to “producing more code per person, closing more sales per rep, and processing more contracts.”
Its AI tools are so productive, in fact, that Anthropic now says human interference is responsible for bottlenecks in productivity. AI agents at the company spend 80% of their session time waiting for human input, rather than actively working.
Though employees agree that using Claude tools improves their productivity, it also shifts their day-to-day responsibilities—and not always for the better.
In an April 2026 survey, 75% of percent of Anthropic employees said they believe the company’s work processes make them productive. Responding to an open-ended question about how AI impacts their daily routines, 20% said their roles have shifted “from actually doing work to managing Claudes,” a task that one employee said is equivalent to babysitting children that can’t accurately gauge the importance of different tasks and problems.
“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.”
Another employee said there’s an “eternal pressure at Anthropic to move as fast as possible and use Claude for everything,” while a third reported that “things we expect in a week are things that would take three months elsewhere.”
When they’re not babysitting agents, the Anthropic staff’s time is also increasingly spent in meetings. The median employee spends 20% more time in meetings compared to a year ago: “A break from managing the agents typically looks like jumping into a meeting,” reads the report, adding that interviewed employees describe “face-to-face coordination as indispensable.”
A more senior team
As more and more work is delegated to AI agents, Anthropic’s hiring needs and practices are shifting to match.
That means making fewer junior hires, with the proportion of junior hires in Anthropic’s technical departments falling from 27% to just 9% since early 2025. The overall share of junior-level technical employees also fell in the same window, shrinking from 21% to 14%.
Anthropic attributes this reduced demand for junior hires to teams offloading work to Claude, including writing code and finding bugs. AI agents at the company now work far longer hours than humans do, with interactive AI agents logging six times as many hours as human employees as of August.
