AI is changing work, and Anthropic studied its own staff to learn exactly how.
Anthropic shared its August findings in a blog post published Tuesday. The study conducted 53 in-depth interviews with 132 engineers and researchers to examine internal use of Anthropic’s agent coding tool, Claude Code. This research aims to understand how AI is transforming work in businesses and society more broadly.
“We find that the use of AI is fundamentally changing the nature of work for software developers, creating both hope and concern,” the blog reads.
The results showed that employees felt more productive and more “full stack,” or able to perform a variety of technical tasks.
For example, the study found that 27% of the work assisted by Claude consisted of tasks that would not have been possible to perform otherwise. These include scaling projects that wouldn’t be cost-effective if done manually, and data dashboards that are nice to have.
Anthropic employees also reported being able to “fully delegate” 0 to 20 percent of their work to Claude, especially “easily verifiable” or “boring” tasks.
But employees also expressed concern about how common AI assistants like Claude are becoming.
“Some noticed that as they collaborated more with AI, they collaborated less with colleagues, and others wondered if they could eventually automate themselves and eliminate their jobs,” the blog said.
Employees said they were concerned about “atrophying the deeper skillsets” needed to write and check code.
“When it becomes so easy and fast to produce output, it becomes increasingly difficult to actually take the time to learn something,” one employee said, according to the report.
Some employees said they missed out on social dynamics and mentoring opportunities.
“Questions that used to go to colleagues are now asked first to Claude,” the report said. One person told investigators: “I like working with people, but it’s sad that I don’t ‘need’ them now… Younger people ask me fewer questions.”
The changes Claude Code brought to internal operations gave software engineers mixed feelings about their future relevance.
“I think I’m optimistic in the short term, but in the long term, AI will eventually do everything and make me and many other people irrelevant,” the blog said, quoting an employee.
Others said it was difficult to predict what their roles would be in a few years.
Beyond Anthropic, employees are showing signs of embracing AI in the workplace and demanding more tools to improve productivity.
According to a January McKinsey report on AI in the workplace, 39% of the 3,613 people surveyed identified themselves as “bloomers,” or AI optimists who want to work with companies to create responsible AI tools. Another 20% believe they want to deploy AI quickly with few guardrails.
McKinsey also found that even employees who reported being skeptical of AI demonstrated familiarity with generative AI tools.
