Everyone Says AI Is Killing Entry-Level Jobs. LinkedIn's New Numbers Tell a Stranger Story

You have heard the story all year: AI is eating the bottom rung of the career ladder, and the class of 2026 is graduating into the worst entry-level market in a generation. It is told at dinner tables, in commencement speeches, and in a hundred LinkedIn posts a day, usually by people several decades past their own first job.

This week LinkedIn itself weighed in, with data instead of vibes. Its economists analyzed hiring patterns across millions of member profiles to find which entry-level roles are actually growing, who is getting hired into them, and what those hires have in common (CNBC). The answer complicates the doom story in a way worth sitting with.

The fastest-growing entry-level job in America is AI engineer, and 2026 is the second year in a row it has held the top spot (CBS News). The same technology being blamed for closing doors on young workers is, by LinkedIn’s count, the single biggest source of new ones.

The scale is not small. Employers added roughly 639,000 AI-related job postings in the US between 2023 and 2025, about 75,000 of them for AI engineers specifically, according to LinkedIn’s analysis (CBS News). “Companies are just gorging on AI talent,” is how Kory Kantenga, LinkedIn’s Head of Economics for the Americas, put it (CBS News). And the people winning these roles skew young: LinkedIn’s data shows Gen Z and millennial members landing the fast-growing, high-paying AI jobs at striking rates, precisely the group the doom story says is being shut out (CNBC).

What does an AI engineer actually do all day? Less science fiction than the title suggests. The role builds and runs AI products, wiring models and agents into the software a business already uses, and the heaviest hiring comes from tech firms and financial services companies (CBS News). It is closer to a plumber for machine intelligence than an inventor of it. That is exactly why the demand is broad: every company installing AI needs someone to connect the pipes, and the banks and insurers doing the installing vastly outnumber the labs doing the inventing.

Two straight years at the top also changes what kind of signal this is. One year could be a hype spike, the kind of thing that shows up in a hiring dataset and vanishes by the next report. A repeat means employers who hired AI engineers in 2025 came back for more in 2026, which is what demand looks like when the first hires paid off. Titles come and go in tech, but budgets that get renewed are the closest thing hiring data has to a confession.

What this means if you’re early in your career

Both stories are true at once, and that is the uncomfortable part. Layoffs across tech are running at record pace this year, a wave we tracked in our report on 2026 passing 2025’s full-year layoff total, and at the same time one category of entry-level work is growing faster than anything LinkedIn measures. The market is not closed. It has moved, and it moved toward people who can work with the technology rather than around it.

Kantenga’s advice to new entrants is blunt: “getting started in the labor market probably does require a certain amount of AI aptitude, or AI literacy, to get ahead” (CBS News). Note what he did not say. Not a PhD, not five years of experience nobody under 25 can have. Literacy. The ability to use these tools on real work and explain what you did.

Here is the satisfying way to hold it all: the ladder’s bottom rung did not disappear, it was unbolted and reattached somewhere else. Two straight years of LinkedIn data now point to the same rung. The graduates having the worst year are the ones still lining up where the ladder used to be; the ones having the best year walked over to where it is now.