The AI jobs data inside Stanford HAI’s 2026 AI Index, released Monday, confirms what many entry-level workers have been experiencing: employment for software developers aged 22 to 25 has fallen nearly 20 percent since late 2022, the exact moment generative AI tools entered mainstream use.
Summary
- The decline is specific to young workers in AI-exposed roles: developers aged 30 and older in the same companies saw employment grow 6 to 12 percent over the same period, while call center hiring dropped 15 percent and similar age-based divergence appeared in accounting, marketing, and customer service.
- The pattern does not appear in occupations with low AI exposure: health aides, production supervisors, and manual laborers saw steady or growing employment across all age groups, confirming the effect is concentrated in roles where AI can replicate the textbook knowledge that early-career workers rely on most heavily.
- Firm surveys cited in the Stanford report indicate executives expect the trend to accelerate, with planned headcount reductions in AI-exposed roles expected to outpace recent cuts, meaning the 20 percent decline in young developer employment may be closer to a starting point than a peak.
As MIT Technology Review noted in its coverage of the index, “the job market is struggling to keep up” with AI development. The Stanford study underpinning the finding used ADP payroll records tracking millions of workers at tens of thousands of companies from 2021 through 2025, one of the largest labor datasets applied to the AI employment question. Researchers, led by Erik Brynjolfsson, were able to rule out alternative explanations including remote work patterns, COVID-era hiring, and broader macroeconomic shifts, leaving the correlation with AI exposure as the most consistent explanation for the divergence between young and older workers in the same roles.
The productivity gains AI is delivering in software development are showing up in the same fields where young employment is contracting. AI can now code for hours at a time and handle basic programming faster and with fewer errors than it could when ChatGPT launched in late 2022.
The mechanism is structural. Young developers enter the workforce with textbook knowledge, the coding syntax and basic algorithms taught in computer science programs. That is precisely what AI tools are best at replicating. Experienced developers carry tacit knowledge, system thinking, and organizational context that AI cannot replicate from a prompt. The result is that AI is not replacing developers in general; it is replacing the entry-level layer that has historically served as the apprenticeship model for the profession. Stanford computer science professor Jan Liphardt described it plainly: graduates are “struggling to find entry-level jobs” in “a dramatic reversal from three years ago.”
What the Pattern Looks Like Across Other Professions
The divergence is not limited to software. Customer service representatives, accountants, and administrative assistants all showed the same age-based split, with workers aged 22 to 25 losing ground while experienced workers in the same companies held steady. Employment for nursing aides and production supervisors, occupations where AI augments rather than replaces human judgment and physical presence, grew faster for young workers than for older ones over the same period.
What This Means for the Labor Market Through 2026
The Stanford index also found that AI is boosting productivity by 14 percent in customer service and 26 percent in software development, and that a third of organizations expect AI to shrink their workforce in the coming year, particularly in service and software. Those sectors are the same ones where the young worker employment decline is already documented, creating a feedback loop between rising AI capability, documented productivity gains, and declining entry-level hiring that the 2026 data shows is already underway rather than still projected.
