Why This Matters: Millions of people are about to lose their jobs — not someday, not eventually, but within the next decade. The experts have run the numbers, compared the models, and their predictions are damning. You need to know what’s coming before it lands on your doorstep.
Over 20 expert predictions collected by AIMultiple’s AI job loss research paint a picture that no corporate talking head wants to hang in their lobby. The numbers range from unsettling to genuinely alarming. Goldman Sachs says 300 million jobs could be automated globally. McKinsey puts up to 375 million workers needing to switch occupational categories by 2030. The World Economic Forum estimates 85 million jobs displaced by 2025 — a year that is already here. And yet, somehow, we’re still having the same tired debate about whether AI is “really” a threat.
It is. Full stop.
Who Gets Hit First
The jobs disappearing fastest aren’t the ones people expect. Yes, factory workers are at risk. But so are data entry clerks, paralegals, junior accountants, customer service reps, and radiologists. White-collar work is not the safe harbor everyone assumed it was.
Automation used to eat physical labor. Now it eats cognitive labor. The moment a machine can read, write, analyze, and respond — which it can — the “get a degree and you’ll be fine” logic collapses completely.
Oxford researchers once estimated 47% of US jobs face high risk of automation. That study is over a decade old. The models have gotten dramatically better since then. The actual percentage today is almost certainly higher.
The Numbers Nobody Wants to Quote at a Career Fair
Here’s what the experts are actually saying, stripped of all the diplomatic cushioning:
- IBM’s CEO said the company would pause hiring for roles AI could replace — roughly 7,800 positions.
- Forrester predicts AI will eliminate 9% of US jobs by 2025, while creating only 2% new ones. That’s a net loss. Do the math.
- Elon Musk — whatever you think of him — has said AI will make most jobs obsolete. He’s not wrong on this one.
- The IMF warned that nearly 40% of global employment is exposed to AI disruption, with advanced economies facing even higher exposure.
These aren’t fringe voices. These are central banks, major consulting firms, and Fortune 500 executives speaking plainly when they think the audience is paying attention.
The Jobs That Will Survive
There are roles AI genuinely struggles with — deep human empathy, physical dexterity in unpredictable environments, original creative vision with real accountability. Nurses. Plumbers. Therapists. Artists who own their voice. These jobs aren’t going anywhere soon.
But here’s the problem: the economy hasn’t built enough of them. And retraining millions of displaced workers into roles requiring years of specialized skill is not a problem you solve with a six-week bootcamp and a LinkedIn badge.
Tech Companies Are Already Doing It
The layoffs happening right now in Silicon Valley are being dressed up as “restructuring” or “efficiency improvements.” They are, in plain language, AI substitution. Why pay a team of five when one model handles the output of all five — around the clock, without benefits, without complaints?
The same companies building tools promising to “empower creators” are quietly reducing their own human headcount. Speaking of creator economies, platforms like Fanvue are showing how AI-powered monetization can actually put money back in human creators’ hands — but they’re the exception, not the rule. Most AI deployment is about reducing labor costs, not expanding human opportunity.
The Hot Take
Universal Basic Income isn’t a radical leftist fantasy anymore — it’s the only policy that actually scales to the size of this problem. Every government pretending retraining programs will absorb the displacement is lying to voters to avoid a hard conversation. When 300 million jobs disappear over 30 years, you cannot retrain your way out of it. At some point, we have to admit the economy will produce enormous wealth AND enormous joblessness simultaneously — and decide who those gains actually belong to. The answer right now is: not the workers who lost the jobs.
What Needs to Happen
Governments need to stop treating AI job loss as a future problem. It’s a present one. Policy needs to move at the speed of the technology — which means now, not after the next election cycle. Corporations need mandatory disclosure of AI-driven workforce reductions. Workers need portable benefits, not ones tied to jobs that may not exist in five years.
The tech industry loves to talk about “the future of work.” Meanwhile, actual workers are trying to figure out if they’ll have work at all. This week, the debate over AI dominance between Meta, OpenAI, Google, and Anthropic is playing out at the CEO level — but the people who should be at that table are the ones whose livelihoods are on the line. They’re not invited. That’s the real story. And it doesn’t end well unless we force it to.