Twelve sources, four continents, one uncomfortable pattern. Here is what the data actually says about who AI is hitting in the job market right now, and what to do about it.
Nobody is coming to save your career. Not your employer, not the government, not the app you are shouting at. That responsibility is yours, and the sooner you accept it, the sooner you stop being a bystander in your own future.
America. Britain. Europe. And it does not stop there. You cannot get experience without a job. You cannot get a job without experience. That trap used to be annoying. Now it is a wall, on three continents at once, and the data just proved it.
Why I am telling you this
I have been saying some version of this since 2014. I sat on a roundtable that year at IBM, where I worked at the time, in a room full of people whose job it was to be excited about the future. One of the executives kept going on and on. AI is incredible. AI is the future. AI is going to change everything. Everyone around the table was nodding along.
I looked at a couple of the people next to me and said what I actually thought. I am genuinely worried about AI and what it is going to do to people’s jobs.
That was over a decade before any of the numbers below existed. I did not need the data to see it coming. Now the data exists, and it says I was right to be worried then. It is also, and I will get to this, why I am not worried anymore.
The stat that started this
Stanford’s Institute for Human-Centered AI released their 2026 AI Index this April. Not a survey, not opinion. ADP payroll records, real employment numbers, across millions of American workers.
Developers aged 22 to 25 are down nearly 20% since 2024. Developers 30 and older, at the exact same companies, over the exact same period, grew. Same firms. Same titles. Same years. One difference: experience, the exact thing the losing group cannot get anymore because the door just closed on them.
The same pattern showed up in customer service. And employers told Stanford directly that a third of them expect more cuts before the year is out.
It is not just America
In the UK, tech graduate jobs fell 46% from 2024, with a further 53% drop projected through 2026, according to the Institute of Student Employers. British government data found postings for the most AI-exposed jobs down 38% since 2022, almost double the drop in everything else. Indeed’s own July numbers put UK graduate postings at their lowest level since 2020, at the exact moment demand for AI skills hit a record high.
Different country, identical story.
It stops being a country problem entirely
In India, a survey of over 1,500 tech workers on the platform Blind found 66% of people working in AI and machine learning itself expect a layoff or a major team cut in the next three to six months. Read that again. The people building the technology are telling researchers they think it is coming for them first.
Worth being honest that this is a self-selected, anonymous survey, not payroll data like Stanford’s, so treat it as what AI workers say they fear rather than a confirmed outcome. But the fear itself is the story, and it cuts against the idea that specialising in AI makes you safe.
The British Standards Institution polled 850 business leaders across Australia, China, France, Germany, Japan, the UK, and the US. Thirty nine percent had already cut entry-level roles because of AI. Forty three percent more said they would before the year was out. That is eighty two percent of leaders across seven economies, already acting or already planning to.
A separate analysis of 126 million job postings worldwide found entry-level roles fell 29 percentage points as a share of all hiring in eighteen months. Junior tech roles specifically were down 35%.
Europe is real but messier
The European Central Bank’s own data this quarter shows youth unemployment pulling away from overall unemployment, a gap that has widened since 2023. But even the ECB will not commit to blaming AI outright. Some of it is simply a slower economy. I am not going to pretend more certainty than actually exists here.
Who to actually be angry at
A number this sharp gets weaponised, so I want to be straight about who deserves the anger. Not the technology.
The executive who fires a 23 year old, tells the press AI made him do it, and collects a bonus for the savings, while his own company’s numbers prove the cut did not even work. Economists at Wharton and Oxford have both called this pattern out by name. A study of 350 companies actively using AI found the ones cutting headcount hardest were not getting better financial results than the ones who cut the least.
And it is not universal. Australia’s graduate postings are actually rebounding this year, and Indeed’s own economists there say the timing does not support AI as the primary cause. A handful of companies are moving the opposite direction entirely. Reddit is hiring more graduates specifically because they call them AI native. IBM expanded entry-level hiring. Dropbox, Cloudflare, and LinkedIn have all grown their graduate programmes.
Hold both facts. The decline across most of the world is real and specific. Not every company blaming AI is telling you the truth about why.
The trap nobody wants to name
Every piece of advice given to 22 to 25 year olds says get more experience. The jobs that used to hand you that experience are the ones disappearing, in country after country, and now even among the people building the AI itself. That is not advice. That is a locked door with a sign telling you to walk through it.
No experience without a job. Increasingly, no job without experience. AI did not build that trap. It just welded the door shut on four continents at once.
What actually works
The bar has not disappeared, it has moved. It used to be: prove yourself by getting hired to do junior work for two years. It is closer now to: prove yourself by showing what you can build, with AI doing the routine part and you directing it.
UK data backs this with a real number. Trainee AI Engineer roles now average nearly £36,000, about 24% above a typical graduate salary, and AI-focused tracks are moving through the pay bands faster than the traditional route.
That means the moves available to you do not require someone handing you a junior seat first. Open source work anyone can check. A portfolio of real, shipped output, even small output. Freelance or contract work that needs no company’s permission to exist. None of that replaces full-time employment. All of it is now proof you can do the work, proof that used to only exist once you already had the job.
You will be okay
Here is what I actually tell people when we talk about this. You will be okay. Not because the data is wrong, it is not, but because the people who come through this fine are not the ones hoping the trend reverses. They are the ones who treat AI fluency, real management and people skills, and unique, hard-to-copy experience as the actual job now, not extras bolted onto the old one.
That combination did not exist as a requirement five years ago. It is the requirement now. Learn to direct AI instead of competing with it. Build the judgment and the people skills that made experienced workers valuable long before any of this started. Go get the kind of experience that cannot be faked or generated, the specific, hard-won stuff that is actually yours.
Do those three things and this entire piece stops being about you.
If you are not 22
If you are reading this and you are not in that 22 to 25 bracket, you are probably the person who actually decides whether someone gets hired, mentored, or let go. You are hiring the graduate, managing the junior on your team, or watching your own kid go through exactly what I have just described. This is not someone else’s problem happening somewhere else. This is the pipeline you are inheriting, whether you built it or not.
Lots of people are worried about their jobs right now, and struggling without the right guidance. If that sounds like you, sign up to Monday Influencer, the link is below.
Want more? Change your life. Subscribe to MONDAY INFLUENCER®
References
United States Stanford Institute for Human-Centered AI, 2026 AI Index, April 2026 (ADP payroll data)
United Kingdom Institute of Student Employers, tech graduate hiring data, 2024-2026 UK Department for Science, Innovation and Technology (DSIT), AI-exposed occupation postings data, 2026 Indeed Hiring Lab UK, graduate postings report, July 2026 UK Trainee AI Engineer salary data, 2026
India Blind, anonymous professional survey, India-based AI/ML workers, July 2026
Global British Standards Institution, survey of 850 business leaders across Australia, China, France, Germany, Japan, UK, US, 2026 Global job postings analysis, 126 million postings, January 2024-mid 2025
Europe European Central Bank, Economic Bulletin, Issue 5/2026
Analysis and Counterpoints Wharton School (Peter Cappelli) and Oxford Economics, commentary on AI-attributed layoffs Study of 350 companies on AI-driven headcount cuts versus financial performance Indeed Hiring Lab Australia, graduate postings rebound data, 2026