Tech companies cut over 45,000 jobs in Q1 2026, and most of them pointed at AI as the reason. Block fired 40% of its workforce. Atlassian dropped 1,600 people. Meta is planning to cut 16,000. But here is the uncomfortable number that nobody is talking about: according to Harvard Business Review, only 2% of companies making these cuts have AI actually performing the work they are eliminating. The other 98% are cutting on speculation.
This is not a story about robots replacing humans. It is a story about executives making bets, sometimes smart, sometimes reckless, on a technology that is still figuring itself out.
The Q1 2026 Numbers: What Actually Happened
The data comes from Challenger, Gray & Christmas, the firm that has tracked U.S. layoffs since the 1990s. Their numbers: 33,330 U.S. tech layoffs in January and February 2026 alone, a 51% increase over the same period in 2025. Globally, RationalFX tracked 45,363 tech job cuts through early March, with about 9,200 of those explicitly linked to AI and automation.
That 20% attribution rate is telling. Companies are simultaneously restructuring for efficiency, cutting middle management, and pivoting product strategies. AI is a convenient umbrella, but the actual picture is messier.
The bigger trend: in 2025, companies attributed 55,000 total layoffs to AI, twelve times the number from the prior year. The acceleration is real even if the framing oversimplifies things.
Who Cut and Why
The biggest moves in Q1 2026:
Block (Square/Cash App) slashed 4,000 jobs, 40% of its workforce. CEO Jack Dorsey was explicit: “intelligence tools” would replace the work. This was the most aggressive AI-driven cut by percentage.
Amazon let go of 16,000 corporate roles in January, framing it as anti-bureaucracy streamlining. AI was not the stated reason, but the company simultaneously ramped AWS AI agent services and hired thousands of ML engineers.
Atlassian cut 1,600 people (10% of staff) in March, citing a “pivot to AI.” Weeks earlier, the company had demoed AI “teammates” handling QA testing and code review.
Meta reportedly plans to cut up to 16,000 roles (20% of workforce) to offset its $135 billion AI infrastructure spend in 2026, according to Reuters.
Salesforce eliminated around 4,000 customer service positions after CEO Marc Benioff stated publicly: “I need less heads with AI.” The company then hired 2,000 salespeople focused on selling its AI agent platform, Agentforce.
Smaller but significant: WiseTech Global (2,000), Crypto.com (180, with the CEO warning about “roles that do not adapt”), and HP (announcing up to 6,000 cuts by 2028).
Which Roles Are Actually at Risk
Not all jobs face the same exposure. Based on data from Axios’s AI disruption tracker and reporting from the Washington Post, the highest-risk roles cluster around structured, repeatable work:
Customer service representatives are ground zero. Klarna replaced 700 support agents with AI in 2025, then quietly started rehiring humans when resolution quality dropped. Salesforce cut 4,000 support roles. The pattern: companies overestimate how much of customer service is “routine.”
Data entry and back-office processing has been automatable since RPA tools launched a decade ago. AI agents accelerate this, but the jobs were already declining.
Software engineering is the surprise entry. Traditional SWE job postings are down while AI-related engineering postings surged 340% since 2024. The role is not disappearing; it is transforming. Companies want engineers who can build and manage AI systems, not just write CRUD endpoints.
Content and copywriting roles took hits at companies like Duolingo (which went “AI-first” and cut contractors) and Chegg (whose entire tutoring model was disrupted by ChatGPT). Quality content still requires humans, but the volume of commodity content work has collapsed.
QA testing is getting absorbed by AI agents. Atlassian’s pre-layoff demos showed AI teammates running test suites. For manual QA, the writing is on the wall.
According to an Index.dev survey, the top reasons companies gave for tech layoffs: AI adoption (44%), reorganization (42%), and budget constraints (39%). Only 9% cited pure performance issues. The takeaway: these cuts are strategic repositioning, not quality-based firing.
The Rehiring Boomerang Nobody Expects
Here is where the narrative gets interesting. Gartner predicts that 50% of companies that cut customer service staff for AI will rehire by 2027. Forrester found that 55% of employers already regret AI-driven layoffs. Forbes ran a headline in March: “Why Today’s AI-Driven Layoffs Are Becoming Tomorrow’s Rehiring Crisis.”
Klarna is the cautionary tale everyone should study. The company loudly replaced 700 support agents with AI, got praised by investors, and then quietly hired humans back when customer satisfaction tanked. The AI handled volume but failed at nuance, escalation, and the kind of empathy that retains high-value customers.
Nvidia CEO Jensen Huang weighed in at GTC 2026 on March 21, telling companies rushing to cut headcount: “You’re out of imagination.” His argument: AI should expand what companies can do, not just shrink their payroll.
The Skills Mismatch Problem
The countertrend data sounds reassuring at first glance. LinkedIn reported at Davos 2026 that 1.3 million new AI-related jobs have been created globally in two years. AI-related job postings are up 340% since 2024. 77% of firms report AI-driven job creation.
But the people being laid off from customer service, QA, and content roles are not the people filling AI engineer positions. A displaced call center agent in Phoenix does not become a machine learning engineer in San Francisco. The jobs being created require fundamentally different skills than the jobs being destroyed, and retraining programs have not scaled to bridge that gap.
Speculation vs. Reality: The 2% Problem
The HBR finding deserves its own section: 60% of companies cutting jobs are doing so in anticipation of AI capabilities, while only 2% report that AI is actually performing the eliminated work today. This means the vast majority of AI layoffs in 2026 are a bet on the future, not a response to present reality.
Some of these bets will pay off. AI agent capabilities are advancing rapidly, and the agent economy is growing. But corporate history is littered with premature restructurings that destroyed institutional knowledge, gutted morale, and left companies scrambling to rebuild when the promised technology did not materialize on schedule.
The smarter companies are taking a different approach. Instead of firing first and automating later, they are redeploying. Deutsche Telekom reskilled 3,000 customer service agents into AI supervision roles. Siemens expanded its workforce by 10,000 while simultaneously deploying AI agents across manufacturing, using humans as oversight layers rather than cutting them.
The difference between companies that will thrive and those that will regret these cuts comes down to a simple question: are you using AI to do things you could not do before, or are you using it as an excuse to do the same things with fewer people?
Frequently Asked Questions
Which companies laid off workers because of AI in 2026?
Major companies cutting jobs and citing AI in Q1 2026 include Block (4,000 jobs, 40% of workforce), Amazon (16,000 corporate roles), Atlassian (1,600, 10% of staff), Meta (planned 16,000), Salesforce (4,000 customer service roles), WiseTech Global (2,000), and Crypto.com (180). Over 45,000 tech jobs were cut globally by March 2026.
Are companies actually replacing workers with AI or is it an excuse?
Harvard Business Review research found that 60% of companies cutting jobs are doing so in anticipation of AI capabilities, but only 2% report that AI is actually performing the eliminated work today. Many analysts describe this as speculative restructuring rather than genuine AI replacement.
Will companies rehire after AI layoffs?
Gartner predicts 50% of companies that cut customer service staff for AI will rehire by 2027. Forrester found that 55% of employers already regret AI-driven layoffs. Klarna is a notable example: after replacing 700 support agents with AI, the company began rehiring humans when customer satisfaction declined.
Which jobs are most at risk from AI automation in 2026?
Customer service representatives, data entry clerks, QA testers, content writers, and traditional software engineers face the highest risk. However, AI-related job postings have surged 340% since 2024, and LinkedIn reports 1.3 million new AI-related jobs created globally. The challenge is a skills mismatch between jobs being cut and jobs being created.
Is AI creating more jobs than it destroys?
LinkedIn data shows 1.3 million new AI-related jobs created globally in two years, and 77% of firms report AI-driven job creation. However, the new roles (AI engineers, data annotators, ML specialists) require fundamentally different skills than the jobs being eliminated, creating a significant retraining gap.