Tech layoffs have hit 45,000 workers in March alone, with over 9,200 attributed directly to AI and automation. But a pattern is emerging that should make executives nervous: more than half the companies that fired workers for AI are now regretting it.
The Numbers Are Staggering
Oracle announced plans to cut 20,000-30,000 employees this week—up to 18% of its global workforce. The reason isn’t efficiency gains from AI. It’s the opposite: Oracle needs to free up $8-10 billion in cash flow to pay for AI data center expansion tied to a $156 billion OpenAI deal requiring 3 million GPUs over five years.
US banks have retreated from financing Oracle’s AI ambitions, doubling borrowing costs and stalling data center projects. The company is now considering selling Cerner to fund its AI bets.
Meanwhile, Amazon has cut roughly 30,000 corporate jobs since October—about 10% of its white-collar workforce. CEO Andy Jassy wrote that “as AI and agents roll out, we will need fewer people doing some of the jobs that are being done today.”
Jack Dorsey’s Block made headlines in February by slashing nearly half its workforce—from 10,000 employees down to under 6,000. Dorsey predicted most companies would make similar cuts within the year. Investors sent Block’s stock up 24%.
The Regret Is Real
Here’s the part those CEOs aren’t tweeting about: according to Forrester’s “Predictions 2026” report, 55% of employers regret AI-related layoffs.
The reasons are damning:
- Nearly a third lost critical skills and expertise when employees left
- 28% said remaining staff couldn’t fill knowledge gaps
- Only about one in five said AI fully replaced eliminated roles without operational issues
Companies aren’t laying off workers because AI has proven it can do their jobs. As Harvard Business Review put it: they’re betting on AI’s potential, not its performance.
The Klarna Case Study
Klarna CEO Sebastian Siemiatkowski learned this lesson publicly. After cutting his workforce from 5,500 to 3,400 and claiming AI chatbots were doing the work of 700 customer service agents, he’s now hiring humans again.
Customer satisfaction had dropped. Service quality became inconsistent. Engineers and marketers were pulled in to help answer customer complaints. Siemiatkowski admitted that AI agents without human support were “not the right fit.”
Forrester predicts half of AI-attributed layoffs will be quietly reversed—but often offshore or at lower salaries.
What Anthropic Found
Anthropic released a new labor market analysis this week that adds nuance to the picture. The company built an “AI exposure index” tracking which jobs are most vulnerable.
The key finding: workers in highly AI-exposed occupations haven’t actually lost jobs at higher rates than those in AI-proof roles. The gap since ChatGPT launched is “small and insignificant.”
But there’s a catch. Anthropic found “suggestive evidence that hiring of younger workers—particularly ages 22 to 25—has slowed in exposed occupations.” AI isn’t firing people. It’s preventing them from getting hired in the first place.
Computer programmers (75% task coverage), customer service reps, and data entry specialists rank among the most exposed occupations.
The Other Side: AI Is Creating Jobs Too
LinkedIn data shows AI has added 1.3 million new roles in two years. Job postings mentioning AI surged 130%. Head of AI positions are proliferating across the US, UK, Germany, and beyond.
Workers with AI skills command a 56% wage premium—more than double the 25% premium from just a year ago. Machine learning engineers average $186,000 annually, with senior roles in San Jose and San Francisco reaching $265,000.
But here’s the gap: the seven million workers now in AI-fluent roles aren’t the same people losing their jobs. Entry-level workers can’t simply become machine learning engineers overnight.
What This Means
The AI job market is splitting into two tracks:
Track 1: The Layoff Treadmill. Companies cut workers, claiming AI will pick up the slack. AI doesn’t. Companies quietly rehire—often abroad, often at lower wages. Repeat.
Track 2: The Skills Premium. Workers who can build, deploy, and manage AI systems command unprecedented salaries. The demand is real. The supply is limited.
If you’re worried about your job, the honest answer is: learn to work with AI tools, but don’t panic about being replaced by AI that doesn’t work yet. The bigger risk is a hiring freeze that prevents you from getting your next role—especially if you’re young.
And if you’re a CEO considering mass layoffs to impress investors with AI transformation talk, consider this: 55% of your peers already regret doing exactly that.
What You Can Do
-
Track actual AI capability, not hype. If your company is laying off workers for AI that “will” do something, that’s speculation, not strategy.
-
Build AI skills strategically. MLOps, prompt engineering, and AI integration are in demand. But so is being the human who understands what the AI can’t do.
-
Document your institutional knowledge. If layoffs come, companies often realize later they need what you knew. Make yourself the person they call back.
-
Watch for the rehiring wave. Forrester says half of AI layoffs will be reversed. If you’re cut, stay connected with former colleagues—the callback may come.