By Ion Anghel · April 2026
In February 2026, Block CEO Jack Dorsey cut his company's headcount nearly in half — from over 10,000 to just under 6,000 employees. The reason, stated plainly in a shareholder letter: "Intelligence tools have changed what it means to build and run a company."
A month later, Atlassian laid off 10% of its global workforce — 1,600 people — citing changes needed for the "AI era." The same quarter, ASML, riding record orders from the AI boom itself, cut 1,700 jobs for "efficiency." Pinterest, Dell, HP, Oracle — the list goes on. Over 100,000 employees were impacted by AI-attributed layoffs in 2025 alone. In the first months of 2026, the number already exceeds 61,000.
There's just one problem: the AI that's supposed to replace these people doesn't exist yet.
The Numbers Behind the Narrative
A Harvard Business Review study published in January 2026, based on a survey of over 1,000 global executives, found something remarkable. Sixty percent of executives admitted they made headcount reductions in anticipation of AI efficiencies. Another 29% reported hiring fewer people than normal. But only 2% said they had made significant staff reductions as a result of actual AI implementation.
Read that again: 60% fired people based on hopes. 2% based on reality.
Forrester's 2026 workforce predictions are even more damning. According to HR Executive, Forrester predicts that half of AI-attributed layoffs will be quietly reversed — with workers rehired offshore at lower salaries. According to their research, 55% of employers already regret laying off workers for AI. The pattern isn't innovation. It's cost arbitrage dressed up as a technology narrative.
JobsPikr's analysis of actual hiring data illustrates this perfectly. Amazon's Seattle job postings dropped from 22,700 in H1 2025 to 4,540 in Q1 2026. Meanwhile, Milan and Pisa emerged as Amazon's second and third largest hiring cities — locations with significantly lower compensation costs. This isn't AI replacing workers. It's workers in expensive cities being replaced by workers in cheaper cities, with AI as the cover story.
The Readiness Paradox
Even if companies genuinely wanted AI to replace these roles today, most couldn't pull it off. Forrester measures what it calls AIQ — artificial intelligence quotient, essentially AI readiness. In 2025, only 16% of workers had high AIQ. That number is projected to reach just 25% by the end of 2026.
The training gap explains why: only 23% of AI decision-makers said their organizations offered prompt engineering training in 2025. Employees are largely teaching themselves through experimentation.
So the situation is this: companies are firing experienced workers who hold institutional knowledge, while simultaneously lacking a trained workforce capable of directing, validating, and quality-controlling the AI systems meant to replace them. They're removing the people who know how the business works before the tools that are supposed to know it instead actually function.
The Klarna Effect
There's a name for what happens next: the Klarna Effect. In early 2024, the buy-now-pay-later company proudly announced that its AI agents were doing the work of 700 customer service representatives. They implemented a hiring freeze. The press coverage was glowing.
By spring 2025, Klarna was hiring again. It turned out that in certain situations, actual humans were required. The company quietly backpedaled from its most aggressive AI replacement claims.
Fortune describes this pattern across industries. Current AI is what researchers call "jagged" — excellent at some tasks, unreliable at others. It can transcribe meeting notes and generate boilerplate code, but it struggles with nuanced reasoning, edge cases, and the kind of contextual judgment that experienced employees bring.
OpenAI CEO Sam Altman himself has used the term "AI washing" — companies blaming unrelated layoffs on AI technology. When even the person who sells the technology says companies are using it as a fig leaf, it's worth paying attention. As an Oxford Economics report from January 2026 suggested, some firms may be dressing up layoffs as a positive narrative rather than admitting to overhiring or underperformance.
The Hidden Costs
The damage goes beyond the people who lose their jobs. Forrester identifies a growing segment of employees they call "coasters" — disengaged workers who don't think their employer deserves their best effort. This group is expected to reach 28% in 2026.
The logic is straightforward. Employees watch colleagues get laid off for AI that never materializes. They see entry-level positions eliminated so new talent can't join the team. They observe offshore arbitrage disguised as technological progress. When a quarter of your remaining workforce is actively withholding discretionary effort, no amount of AI is going to compensate for that productivity loss.
There's also a generational irony. According to Forrester, Gen Z workers have the highest AI readiness at 22%, compared to just 6% for Baby Boomers. Yet companies are disproportionately eliminating the entry-level positions that Gen Z needs to enter the workforce. The cohort most capable of working with AI is being shut out of the job market by the AI narrative.
What Should Actually Happen
AI will reshape how we work. That much is clear. But the companies doing it right aren't making headlines with mass layoffs. They're doing something much less dramatic: incrementally redesigning workflows, investing in training, and measuring actual productivity gains before making staffing decisions.
The difference between a good AI strategy and a bad one isn't the technology. It's whether you're making decisions based on data or on investor expectations. Firing people because you believe AI will eventually replace them is like selling your car because you heard teleportation might work someday.
Companies should focus on where AI actually delivers value today, train their workforce to use these tools effectively, and make staffing changes only when — and only when — they have evidence that the technology works in their specific context.
AI's impact on work will be real. But it should be measured, gradual, and honest — not a press release.
Disclaimer: I believe AI is the future and that its impact on productivity will be substantial. But "the future" is not an excuse for premature decisions that hurt real people today. The companies that will lead the AI transition are the ones doing it responsibly — with evidence, not hype.