AI company Anthropic has reported a significant decline in junior technical hiring as its employees increasingly use Claude AI agents to automate coding, sales, and administrative tasks. The proportion of junior employees in technical departments fell from 21% to 14% since early 2025, while junior technical hires dropped from 27% to 9%.
According to Anthropic’s workforce research report, published October 16, 2026, the changes occurred alongside rising employee productivity and a shift toward managing AI-generated work. The company examined internal employment records, AI usage, productivity measurements, hiring decisions, employee surveys, and interviews to understand how automation is changing its operations.
Anthropic Reduces Junior Hiring While Expanding Its Workforce
Anthropic found that technical departments now account for 56% of its workforce, down from 61% a year earlier. Technical hiring continues to grow in absolute terms, but recruitment in areas such as sales, finance, and other business functions has increased more rapidly.
The company attributed part of this change to business expansion and the growing ability of non-technical employees to develop software using AI coding tools. Among non-technical workers, monthly coding-agent sessions increased from two in August 2025 to 90 in May 2026.
Hiring managers are also placing greater emphasis on candidates’ independence, judgment, and ability to work with AI systems. Specific technical knowledge and implementation experience have become relatively less prominent in some hiring assessments.
AI Agents Increase Productivity Across Departments
By August 2026, interactive AI agents were performing approximately six times as many working hours as Anthropic’s employees, based on the company’s measurement methods. AI-generated coding activity expanded considerably, with code commits per engineer increasing fivefold over six months.
Productivity improvements extended beyond engineering, with sales employees closing more deals and legal teams processing more contracts. Anthropic also reported that its code-quality measurements showed no clear deterioration in AI-generated work, although the company acknowledged that producing more code does not necessarily translate into greater business value.
In an April 2026 employee survey, 27% of respondents described productivity improvements from Claude, while 20% said their responsibilities had shifted toward supervising AI agents. Employees increasingly reported spending time reviewing AI output, coordinating automated workflows, and developing tools to monitor agent performance.
Human Attention Becomes a Productivity Bottleneck
Despite automation gains, Anthropic found that human attention and judgment are increasingly limiting how quickly AI-generated work can be processed. Its research estimated that reading all the output produced by interactive Claude sessions could occupy more than half the working week for some non-technical employees.
AI agents also spent approximately 80% of their interactive session time waiting for human input. Employees responded by developing monitoring dashboards, automated review systems, and workflows that allow agents to coordinate with one another.
However, automation has not reduced the need for human collaboration. Anthropic’s calendar data showed that median meeting time per employee increased 20% over the previous year, including a 16% increase among technical individual contributors.
The company emphasized that its experience may not represent how AI adoption will develop across the broader economy. Nevertheless, its findings suggest that AI-intensive workplaces may increasingly prioritize employees who can supervise automated systems, make decisions, and coordinate complex work rather than perform every task directly.
Featured image credits: Wikimedia Commons
For more stories like it, click the +Follow button at the top of this page to follow us.