Entry-level job postings in the United States dropped 35% between January 2023 and early 2025, according to data from Revelio Labs. That's not a blip from a single bad quarter. That's a structural shift, and it's accelerating.
Most of the conversation around artificial intelligence and jobs focuses on the wrong time horizon. People debate whether AI will take their current role in the next two years. The more important question is what happens to the pipeline in the next 25.
The Numbers Are Already Stark
Entry-level tech hiring fell 25% year-over-year in 2024. A Stanford study published in 2025 found that workers aged 22 to 25 in AI-sensitive jobs saw a 13% employment reduction since late 2022. SignalFire tracked new job opportunities for people with less than one year of post-graduate experience from 2019 to 2024 and found a 50% drop.
It's not just hiring slowdowns. A 2024 survey found that 37% of employers would rather deploy AI than hire a recent graduate for certain tasks. And 60% of employers who did hire new graduates fired them within a year. Companies aren't just hiring fewer entry-level workers. Many are deciding the economics don't work at all.
Dario Amodei, the CEO of Anthropic, has said he thinks 50% of entry-level roles could be eliminated by AI. Two-fifths of global business leaders surveyed in 2025 said they've already reduced or cut entry-level positions because of AI efficiencies. At Davos 2026, International Monetary Fund Managing Director Kristalina Georgieva described AI's effect on entry-level employment as "a tsunami hitting the labor market."
This Isn't Just a Jobs Problem
Here's what I think gets missed in the coverage. The loss of junior roles isn't only an economic problem for young people today. It's a leadership development problem for every organization in 2040 and beyond.
Senior roles require judgment. They require knowing when the data is telling the wrong story, how to manage a difficult stakeholder, when to push back and when to let something go. Those skills don't come from reading about them. They come from doing junior work: making small decisions with real consequences, getting things wrong in low-stakes environments, learning to read a room over years of practice.
There's no shortcut to that. You can't download context. You can't skip the years of pattern recognition that come from working alongside experienced people and watching how they handle things. The senior leader of 2051 is supposed to be the junior hire of today. If that person never gets hired, who fills the seat in 25 years?
What AI Can't Teach
AI is genuinely good at execution. It can draft, summarize, analyze, and synthesize faster than any junior analyst. That's exactly why companies are making the trade. The short-term math makes sense.
But Forbes put it clearly: AI excels at execution but struggles with recognizing when something is amiss or understanding the reasons behind it. That's not a knock on the technology. It's a description of what the technology is. It does what it's pointed at. It doesn't know when the question itself is the wrong one.
The judgment to know when something is off, the instinct that tells you a stakeholder's silence means something, the experience to recognize a pattern before it becomes a crisis: those things develop through years of doing work where the stakes are real, the feedback is human, and the environment is messy. That's junior work. And we're eliminating it.
The Burnout Hiding in Plain Sight
There's a more immediate problem too. Companies that have already thinned their junior layers are watching their senior staff burn out. The work didn't disappear when the entry-level headcount did. The work got absorbed upward.
And when senior people leave those organizations, there's often no urgency to replace them, because the assumption is that AI will cover it. So the institutional knowledge walks out the door with them, and what's left is a layer of tools with nobody experienced enough to know when the tools are wrong.
Community College Daily noted that as pathways to advancement become less linear, serious questions arise about how future leaders will actually be developed. That's a measured way of describing what is, in practice, a broken pipeline.
The 25-Year Question
I'm not arguing that companies should ignore efficiency. I understand why the economics of hiring a junior analyst look different in 2026 than they did in 2019. The cost-benefit calculation changed, and organizations are responding rationally to the incentives in front of them.
But rational short-term decisions can produce irrational long-term outcomes. We're making a collective choice, company by company, to skip a generation of development. The assumption seems to be that we'll figure out the leadership pipeline problem later, or that the next generation will find a different path to expertise.
Maybe that's true. Maybe new apprenticeship models emerge. Maybe AI tools accelerate learning in ways that compensate for the lost experience. I genuinely don't know.
What I do know is that nobody's really asking the question yet. And by the time the answer matters, the people who should be senior leaders in 2051 will be in their late 40s, looking back at a decade when the door was closed before they ever had a chance to walk through it.
So here's the question I keep coming back to: if we've decided that AI can do the junior work, what's our actual plan for teaching the next generation what it takes to do the senior work?