The defining feature of the 2026 labour market is not the laid off veteran but the graduate who never gets hired in the first place. Stanford's 2026 AI Index found that employment for software developers aged 22 to 25 has fallen by almost 20 percent since 2024, even as employment for developers aged 30 and over grew by 6 to 12 percent over the same period. The door is closing at the bottom of the ladder while the higher rungs hold firm.
A statistical squeeze, not yet a collapse
The aggregate numbers are large but ambiguous. The layoffs tracker TrueUp counted 148,092 tech workers displaced in the first five months of 2026, a daily rate of roughly 981 that runs about 46 percent above the 2025 average, with a full year projection near 370,000. Goldman Sachs estimates that around 16,000 United States jobs a month are now affected by AI. Yet analysts caution that productivity gains remain hard to prove, leaving the data unable to either confirm or refute an AI unemployment shock.
The tasks juniors did are the tasks AI does first
The mechanism is specific. Boilerplate code, scripted testing and routine bug fixing are exactly the assignments that once trained junior developers, and they are also the easiest work to automate. Openings for machine learning engineers now sit about 59 percent above their pre pandemic baseline while general software engineering openings remain roughly 49 percent below it. The work has not vanished so much as moved up the skill curve, out of reach of the newcomer.
Employers are rewriting the entry rung
That shift is visible in hiring plans. The share of companies intending to reduce junior roles has jumped from about 17 percent to 43 percent in a single year, and roughly a third say they are tilting their workforce mix toward mid level staff. According to NACE, some 35 percent of early career postings now ask for AI skills, nearly triple the share of a year earlier. The job has not disappeared, but its description has changed faster than most degree courses can follow.
A few firms are betting the other way
Not everyone is pulling up the ladder. IBM says it is tripling its United States entry level hiring in 2026, and Salesforce is targeting 1,000 AI native graduates to staff its Agentforce platform, wagering that young workers fluent in AI from the start are an asset rather than a cost. For this year's graduates the lesson is stark. The first job will look nothing like the one their predecessors held, and the ticket in is demonstrable fluency with the very tools now doing the old entry level work.
