In a report published October 16, 2026, Anthropic said junior candidates’ share of technical hires had fallen from 27% to 9% since early 2025, while the company’s total technical hiring continued to grow in absolute terms. The employment report, based on internal company data and 17 employee interviews, describes changes at Anthropic.
What the two junior-hiring figures measure
The 27%-to-9% figure concerns new technical hires; the 21%-to-14% figure concerns employees already in technical departments. Both changes cover the period since early 2025, but the percentages have different denominators.
| Measure | What the percentage describes | Reported change | Period |
| Junior share of technical hires | New technical hires at junior levels | 27% to 9% | Since early 2025 |
| Junior share of technical employees | Employees in Anthropic’s technical departments | 21% to 14% | Since early 2025 |
Those falling shares do not mean technical hiring stopped. Anthropic said absolute technical hiring continued to increase, even as technical departments’ share of all employees fell from 61% to 56% over the prior year. Hiring in sales, finance, and recruiting grew faster.
What changed in candidate assessments
Anthropic said hiring managers increasingly emphasized autonomy and judgment, while placing less emphasis on domain expertise and implementation. Mentions of AI-tool use in technical hiring assessments rose from about 5% in 2024 to three times that rate in 2026. In nontechnical assessments, mentions rose from 7% in 2024 to four times that rate.
In an April 2026 manager survey, around one-third of machine-learning managers and one-fifth of software-engineering managers said AI had led them to hire less. Those responses describe managers’ views; Anthropic did not estimate a causal effect of AI on junior hiring.
How the report connects AI use and hiring
Anthropic reported that code commits per engineer reached five times their level six months earlier. The company also said employee-directed interactive agents logged six times the work hours of their human handlers by August 2026. Its measure counted agent runtime, including inference and active tool use, while excluding waits longer than 30 minutes; human hours were estimated at 40 hours per week per employee. Commit volume alone does not measure code quality or business impact.
For a separate coding-agent analysis, Anthropic compared 192 engineers who adopted coding agents with 74 who did not. The company counted adoption from the first month with at least five agentic sessions. Adoption was not random: engineers who later adopted fully autonomous agents were already on a steeply rising output trajectory. These productivity findings do not establish that AI caused the change in junior hiring.
What Anthropic proposed for current employees
Anthropic reported that employees who had worked there for at least a year were 50% more likely to move between departments each month in 2026 than in 2025. The company proposed internal redeployment and an early-warning system for roles facing change. Anthropic said those measures for current employees do not directly address the slowdown in junior hiring, and that external hiring can become a bottleneck as work changes.
