Is AI squeezing entry-level tech jobs? Anthropic’s junior technical hiring drops from 27% to 9%

Artificial Intelligence (AI) is gradually shifting how companies prioritise hiring. While the IT sector is heavily investing in entry-level jobs, Anthropic reports a drop in junior technical hires from 27% to 9% since December 2025, showcasing how AI adoption could reshape hiring priorities and reduce opportunities for entry-level technical roles across the industry.

According to The Anthropic Institute report, titled How Employment is Changing at Anthropic,  Junior employees accounted for 21% of Anthropic’s technical workforce in early 2025. That share has since fallen to 14%. This does not mean that Anthropic is laying off junior-level employees. Its overall workforce is increasing, but junior employees make up a smaller proportion of technical teams.

Must read: Amazon plans potential 5,000 job cuts, 15 warehouse closures in New York: Report

It was highlighted that software engineers deploy the majority of entry-level tasks or routine development to AI; therefore, in the coming months, companies may need fewer beginners to complete tasks.

The study further revealed that AI is also changing the roles of people already working in tech teams. Managers and other non-engineering employees are increasingly involved in writing and deploying code. The number of code commits made by managers has also doubled. Therefore, rather than just supervising teams, coordinating projects, or planning products, managers have also started to contribute directly to software development.

AI’s productivity gains at Anthropic

The report highlighted that the productivity gains of AI at Anthropic are significant, measurable, and span both technical and non-technical roles. With greater AI adoption, output has risen across various business units. Lawyers are processing more contracts, accounts payable is processing more invoices, and salespeople are closing more deals.

Must read: Anthropic commits 5% of its shares to public benefit fund: What it means & how it will work

“In March 2026, the sum of autonomous and interactive agent work hours leapt past human work hours for the first time,” the company said. By August, individual employee-directed interactive agents were working 6 times as many hours as their human handlers.

However, the study highlighted that human attention is becoming a major limitation as employees spend more time managing and monitoring AI tools rather than doing the work directly. AI agents spend around 80% of their session time waiting for human input. This means they often pause for instructions, decisions, feedback, or approvals.

This further revealed that meta-work that involves setting up, directing, and supervising AI systems increased from 10% of interactive AI sessions in autumn 2025 to 22% in spring 2026. Therefore, AI does not necessarily eliminate work; it can also create new responsibilities.