Oracle fired up to 30,000 people on Monday with a 6 a.m. email. Meta is rumored to be next, with up to 15,000 cuts possibly landing as early as April 8. And the latest federal data shows U.S. hiring has cratered to its weakest rate since the pandemic lockdowns.
Welcome to Q2 2026. Here’s what’s happening in the AI job market.
Oracle: 30,000 Gone to Feed the AI Machine
Oracle began executing layoffs on March 31, notifying employees across the U.S., India, Canada, and Mexico via a cold “Oracle Leadership” email that arrived at roughly 6 a.m. local time. No prior warning from HR. No heads-up from managers.
Investment bank TD Cowen estimates the cuts will hit 20,000 to 30,000 workers—roughly 18% of Oracle’s 162,000-person global workforce. The freed-up cash, an estimated $8–10 billion, goes straight into Oracle Cloud Infrastructure’s AI data center buildout.
The math tells the story. Oracle posted a 95% jump in net income last quarter ($6.13 billion) and has $523 billion in remaining performance obligations—up 433% year-over-year. The company isn’t struggling. It’s choosing to trade human workers for GPU clusters.
Oracle has also taken on $58 billion in new debt within two months to fund infrastructure. The employees were, apparently, the cheaper line item to cut.
Meta: 15,000 More Cuts Looming
Meta is next in line. Reuters reported in mid-March that senior executives were directed to plan workforce reductions of roughly 20%—about 15,000 positions from a workforce of 79,000.
The company has already started trimming. Hundreds of employees were cut across five divisions including Facebook, Reality Labs, global operations, recruiting, and sales. Employee chatter on Blind points to April 8 as the date for the larger wave, though Meta called the reports “speculative.”
What’s not speculative: Meta plans to spend $135 billion on AI in 2026, more than doubling last year’s budget. Wall Street loves it—Meta’s stock climbed nearly 3% on the layoff news. The market has decided that firing humans to fund AI is a buy signal.
A leaked internal memo revealed plans to replace “human-led processes with automated Llama-based systems,” with cuts targeted at mid-level management and non-technical departments.
The Macro Picture Is Ugly
The Bureau of Labor Statistics’ February jobs report showed the U.S. economy shed 92,000 jobs—well below the 59,000 additions that economists had expected.
Then the JOLTS report dropped on March 31. U.S. businesses added just 4.85 million workers in February—the fewest since April 2020. The hiring rate fell to 3.1% of total employment, the lowest since the pandemic lockdowns and, before that, 2011.
Job openings slipped to 6.88 million. Voluntary quits fell to 2.97 million, the lowest since 2020. Economists call it a “low-hire, low-fire” market: companies won’t add staff but won’t let go of the people they have either. Unless, of course, they can blame AI.
The March jobs report drops April 3. Expectations are grim.
Q1 2026 by the Numbers
The quarter’s final tally is staggering:
- ~60,000 confirmed tech job cuts across 200+ companies, up 15% from two weeks prior
- 23% of Q1 layoffs explicitly cite AI automation in SEC filings—up from 14% in Q4 2025
- Block: 4,000 jobs (40% of workforce), CEO Jack Dorsey called it an AI inevitability
- Atlassian: 1,600 jobs (10% of workforce), simultaneously hiring 800 AI-focused roles
- Oracle: Up to 30,000 jobs to fund AI data centers
- Meta: 700+ already cut, up to 15,000 more expected
- Epic Games: 1,000 jobs (notably, not blaming AI)
The Atlassian model is becoming the template: cut traditional roles in content creation, customer support, QA, and project management, then hire a smaller number of AI engineers. Net headcount drops. The workforce “reshapes.”
Harvard Study: AI Restructures, Not Just Eliminates
A Harvard Business School study published in HBR analyzed nearly all U.S. job vacancies from 2019 through early 2025. The findings are more nuanced than the layoff headlines suggest.
AI is cutting 17% of job postings in automation-heavy roles—but increasing demand by 22% in positions that benefit from human-AI collaboration. The study found that employers aren’t just eliminating jobs; they’re fundamentally restructuring what those roles involve.
The most exposed roles: data entry, basic financial analysis, standard report generation, routine compliance monitoring, junior software development, QA testing, entry-level IT support, and administrative scheduling.
In a companion survey of 2,357 people across 940 occupations, 94% said they prefer AI being used as a collaborative tool rather than a full replacement. The workforce has an opinion. It’s just not the one being asked for.
Where the Money Is
If you’re on the job market, the demand side tells a clear story. AI-specific roles grew 81% year-over-year, though they still represent a small share of total tech hiring. Over 275,000 active U.S. job postings in January referenced AI skills.
The skills commanding the highest premiums:
- LLM fine-tuning: 25–40% salary premium over generalist ML engineers
- MLOps: 20–35% premium
- Prompt engineering: demand surged 135% this year
- RAG architecture, LangChain, PyTorch: the three most-requested skills on LinkedIn AI listings
Average AI engineer base pay sits between $140,000 and $185,000, with senior total compensation regularly clearing $300,000. The talent gap remains severe: 72% of employers globally say they can’t find the AI skills they need, with demand outstripping supply roughly 3-to-1.
What This Means
We’re watching a market split in real time. On one side: mass layoffs driven by AI spending, an executive class that treats workforce reduction as a feature rather than a bug, and a broader economy where overall hiring has fallen off a cliff. On the other: a white-hot market for anyone who can build, deploy, or govern AI systems.
The uncomfortable truth is that both sides are true simultaneously. Oracle can fire 30,000 people while complaining about a talent shortage. Atlassian can cut 1,600 roles and open 800 new ones in the same press release. The jobs aren’t disappearing—they’re being replaced by different, more specialized jobs that fewer people are qualified for.
What You Can Do
If you’re employed in a role that touches AI: Start documenting how you use AI tools to augment your work. The Harvard study shows that hybrid human-AI roles are growing. Position yourself in that category.
If you’re job hunting: LLM fine-tuning, RAG architecture, and MLOps are where the money is. Free resources exist—Hugging Face courses, fast.ai, and LangChain documentation are good starting points. The 3-to-1 demand-supply ratio means employers are willing to train, but you need baseline competency.
If you’re in a “restructuring” role: Customer support, QA, content creation, and administrative positions are being automated first. Lateral moves into AI-adjacent roles (AI training data curation, prompt engineering, AI compliance) offer paths that don’t require starting from scratch.
If you just want to understand the real numbers: The March jobs report drops April 3. Watch it closely. If the trend continues, we may be looking at the first AI-attributed recession in hiring—not because the economy is contracting, but because companies are choosing to spend on silicon instead of salaries.