In a report released this Thursday, Anthropic asserts that its internal adoption of AI agents has fundamentally altered its operations, leading to increased productivity and a shift in labor demand. The company claims that it is hiring fewer junior technical staff while human oversight has become the primary bottleneck in its workflows. As of May 2026, the report states that AI agents officially logged more work hours than human employees. Employees are now reportedly dedicating 22% of their time to managing Claude agents, marking a significant increase from the previous year. Neither of those predictions came to pass. According to anthropic, AI has changed the way work gets done, and as a result, restructured employment at the company. Software engineers could face a tougher job market while entry-level tech workers struggle to build necessary skills and experience, the report suggests its changes foreshadow the future of work more broadly: as more companies adopt AI agents. According to by outsourcing some coding and sales-related tasks to AI agents, Anthropic, It raised productivity within a few months. More than double the amount a year ago, based on Anthropic’s analysis of their chat summaries, employees now spend about 22% of their time managing Claude agents. And described its workers as merely “a thin layer atop a growing contingent of AI agents”, it attributes the shifts to AI usage. Some AI researchers are skeptical that it will herald big changes across companies or industries, while the report provides a small glimpse of the inner workings of the AI heavyweight. Months before ChatGPT’s release, according to the paper, state unemployment claims reveal that job markets tanked starting at the beginning of 2022. “It doesn’t appear that the launch of ChatGPT corresponds to any major increase in loss or worsening outcomes,” he said. Who described facing “an army of Claudes, with every one asking for my attention”, anthropic’s report quotes an unnamed engineer. A majority of Americans are wary of AI, with particular concerns about AI-related job losses, according to Pew Research Center, as generative AI products proliferate. Software engineers and other tech workers will surely be affected, Baker said, regardless of when or whether the bubble pops.
The company maintains that this deep integration of AI allows for higher output, including more code per person, increased sales conversion rates, and more efficient contract processing. Reflecting on future industry trends, Anthropic envisions a model of single-person startups where a sole human founder manages swarms of AI agents to generate significant revenue. However, these assertions of efficiency are being met with doubt by outside researchers who suggest the document may be designed to bolster the firm’s valuation ahead of its upcoming initial public offering (IPO). Instead, they say the findings are probably Anthropic’s attempt to drive its valuation ahead of its initial public offering – similar to how SpaceX boasted about proposed space datacenters to boost its market debut in June. “My hope is that there is a reckoning, a reassessment or re-evaluation of how much this has always been about human work,” they said.
Critics, including Yale social scientist Julián Posada, point out that the report functions more like a marketing tool for potential investors than an objective economic study.
Posada noted the lack of independent audit, stating the findings omit crucial context regarding employee sentiment and the actual nature of human-AI engagement. Similarly, Dylan Baker of the Distributed AI Research Institute questioned the sustainability of Anthropic’s strategy. Baker criticized the characterization of human attention as a “bottleneck,” arguing that software development is fundamentally an interpersonal and social process that cannot be fully delegated to AI agents without compromising design quality.
Dario Amodei, Anthropic’s CEO, previously predicted in 2025 that AI would replace half of entry-level white-collar roles and handle 90% of code generation within six months—forecasts that did not materialize.
The recent report claims that since March 2025, entry-level technical hiring dropped from roughly 30% to 9%, while total headcount growth slowed to 14% from 21%.
Researchers like Morgan Frank of Case Western Reserve University warn against drawing broad conclusions from such corporate data.
Frank’s own research indicates that job markets exposed to AI experienced declines well before the debut of tools like ChatGPT, suggesting that current layoffs may be influenced by broader factors or “AI-washing,” where companies attribute standard restructuring to technological progress.
These operational claims arrive at a precarious moment for the AI developer. A draft of the company’s S-1 filing indicates an $80 billion net loss in the first half of the year, alongside a junk credit rating assigned in June. While original market estimates once reached $2 trillion, investors have since downgraded valuations to approximately $1.5 trillion due to slowing revenue growth. The company’s cloud partner, CoreWeave, remains a significant factor in its financial ecosystem as the firm prepares for a possible stock market debut as early as November.
To address growing public anxiety surrounding AI-driven unemployment, Anthropic suggested in its report that firms should grant employees stock ownership as part of a pro-worker adoption strategy. This follows the company’s recent pledge to dedicate 5% of its shares to “public good” initiatives, a move that skeptics have compared to traditional corporate reputation-building tactics. Despite the breadth of the data presented, Anthropic did not provide a comment in response to inquiries regarding its findings or its financial outlook. According to “It is written from the perspective of a company that wants to sell shares, a company that really wants people to invest in it,”, Posada, author of the book Platform Extractivism: Data Work and the People Powering Artificial Intelligence. That doesn’t sound like a fruitful work environment, Baker said. “I wouldn’t think really thoughtful, well-designed products are coming out of that team.”. Critics noted that philanthropy has long doubled as a reputation builder used by tech titans like Bill Gates and Mark Zuckerberg.
