Anthropic Cuts Junior Technical Hiring as AI Agents Take on More Work

Anthropic has reduced the share of junior technical employees in its workforce from 21% to 14% since early 2025, while the proportion of junior technical hires has fallen from 27% to 9%, as the artificial intelligence company increasingly relies on AI agents to perform coding and other tasks, according to a new report published recently by the AI company.

The findings point to a potential shift in entry-level employment as companies integrate AI into their operations. While Anthropic's overall workforce continues to grow, hiring in technical roles has slowed relative to other functions, raising questions about how AI adoption could affect opportunities for workers entering the technology industry.

The report, How Employment Is Changing at Anthropic, draws on internal employment records, AI usage data, productivity measures, employee surveys and interviews with 17 employees.

AI Agents Change Hiring Priorities

Anthropic's findings suggest that AI is changing not only the volume of work employees can complete but also the skills employers prioritise when hiring.

The company reported that, by August 2026, interactive AI agent hours outnumbered employee hours by six to one. The measure includes agent runtime and active tool use, excluding waiting periods longer than 30 minutes. It does not mean AI agents independently completed six times as much work as employees.

According to the report, engineering and research teams began discussing reduced demand for junior technical employees in January 2026. By April, these teams had placed less emphasis on hiring for lower-level technical positions, as coding agents increasingly handled tasks such as writing code and identifying bugs.

In an April survey cited in the report, around one-third of machine learning managers and one-fifth of software engineering managers said they were hiring less because of AI.

Hiring managers are also placing greater emphasis on candidates' ability to use AI tools, exercise independent judgment and work autonomously. The report found that references to AI tool use in technical hiring assessments tripled between 2024 and 2026, while they quadrupled for non-technical roles.

Productivity Gains Extend Beyond Engineering

The changes are not confined to software development. Anthropic found that employees across sales, legal, finance and other functions were using AI agents to automate routine tasks and increase output.

The company reported that code commits per engineer had risen to five times their level six months earlier. Lawyers were processing more contracts, accounts payable teams were handling invoices, and sales representatives were closing more deals.

To examine the relationship between AI adoption and productivity, the researchers tracked 192 engineers who adopted coding agents and compared them with 74 who did not adopt the tools during the period studied. They found that the volume of code added increased sharply after adoption.

However, the authors cautioned that adoption was not random and that some employees may have taken up AI tools while already working on larger projects. They also noted that more code does not necessarily translate into better software or greater business impact.

As AI handles more implementation work, employees are increasingly responsible for reviewing outputs, managing agents and making decisions that require organisational context. The report found that the share of interactive Claude sessions devoted to this kind of supporting work rose from 10% in autumn 2025 to 22% in spring 2026.

Entry-Level Employment Remains a Concern

Anthropic said it has established an early-warning system to identify roles facing potential disruption and has prioritised moving existing employees into other functions where their skills remain relevant. Employees who had been at the company for at least a year were 50% more likely to move between departments each month in 2026 than in 2025.

The authors warned that redeployment policies for existing employees do not address the opportunities lost when companies stop hiring entry-level workers. They said employer initiatives would need to be accompanied by government policies on retraining and income support.

The report also draws on earlier research suggesting that AI-native companies and businesses adopting AI may operate with flatter organisational structures and fewer entry-level employees.

However, Anthropic's experience is not necessarily representative of the wider labour market. The company is an AI developer with unusually high adoption of its own technology, and its employment patterns cannot establish that AI is already causing broad-based job losses.

The findings instead offer an early indication of how work could be reorganised as AI agents become more capable. For technology employers, the challenge will be to capture productivity gains while preserving pathways through which new workers can acquire skills and experience.