AI Tools Reshape Software Hiring as Entry-Level Developer Jobs Fall 20% in the US

Employment of young software developers in the United States has fallen by 20% since ChatGPT was launched in November 2022, according to data highlighted by The Kobeissi Letter and drawn from the Canaries Dashboard of the Stanford Digital Economy Lab, which uses payroll data from ADP Research. Over the same period, employment among developers aged 41 to 49 rose by 20%, opening a gap of roughly 40 percentage points between the youngest and oldest groups of workers.

The decline is concentrated among developers aged 22 to 25, the group that includes most fresh technology graduates. Employment for developers aged 26 to 30 slipped by 4%, while workers in the 31 to 34 and 35 to 40 brackets also recorded growth. Researchers rank software development among the occupations most exposed to artificial intelligence.

The likely explanation is that AI coding tools now handle much of the routine work once assigned to junior developers, including writing basic code, fixing bugs and testing. Companies can complete that work with fewer new hires, while experienced developers who can review and guide AI-generated output remain in demand.

Stanford researchers caution that the figures show a correlation rather than proof that AI alone caused the decline. The dataset also covers only firms that use ADP payroll services, not the entire US labour market, so the findings should be read as a strong signal rather than a complete picture.

For new computer science graduates, the takeaway is that a degree on its own may no longer be sufficient. Employers increasingly look for candidates who can work effectively with AI tools and arrive with practical project experience from day one.

The trend echoes wider debate about how generative AI is reshaping white-collar work, with technology firms adjusting entry-level roles while continuing to compete for experienced engineers who can supervise automated systems.