Research from the Stanford Digital Economy Lab found that software developer employment among workers aged 22 to 25 fell nearly 20% by September 2025 compared to late-2022 levels, right around the time OpenAI’s chatbot went mainstream. For young developers in roles most exposed to AI automation, the relative decline hit 16%.

Before ChatGPT launched in November 2022, programming-heavy jobs were growing at an annual rate of nearly 5%.

The numbers paint a stark picture

A Federal Reserve report released in April 2026 quantified the damage more broadly. Job growth in US programming roles nearly halved after ChatGPT’s launch, translating to roughly 500,000 fewer jobs expected over a three-year period.

The year-over-year contraction for 22-to-25-year-olds in AI-exposed occupations ran at 4.2%, compared to just 1.7% in less-exposed fields.

The Stanford researchers framed this as an “occupation-specific shock” rather than a general industry downturn. Senior engineers, systems architects, and machine learning specialists weren’t getting axed at the same rate. The pain concentrated at the bottom of the career ladder, precisely where young workers need to gain experience before climbing it.

The counterargument is real, but limited

A June 2026 analysis from Ramp and Revelio Labs, covering more than 21,500 US companies, found that firms with higher AI spending actually expanded their workforces by approximately 10%. Even more surprising: entry-level hiring at those high-AI-spending companies grew by around 12% post-adoption.

What it suggests is a bifurcation in the tech labor market. Companies that invested heavily in AI tools didn’t just use them to replace workers. They used them to scale operations, which required more people, including junior ones, to manage, implement, and build on top of those AI systems.

Based on the aggregate numbers, the net effect remains deeply negative for the 22-to-25 cohort.