14,000+ Tech Workers Cut as Oracle, Meta, and Block Bet Big on AI

Oracle just cut 10,000 jobs. Meta eliminated another 3,600. Block shed 1,000 more. And we're barely three months into 2026.

These aren't random cost cuts. Each company pointed to the same culprit: AI can now do the work faster, cheaper, and (they claim) better. Oracle's CFO literally said their AI agents are handling customer support tickets that used to require full teams. Meta's automation push means fewer content moderators and data labelers. Block's AI fraud detection replaced an entire department.

But here's what nobody's saying out loud. These companies aren't struggling. Oracle's cloud revenue is up 23% year-over-year. Meta's ad business is printing money. They're cutting because they can, not because they have to.

<h2>The Numbers Don't Lie</h2>

Tech layoffs hit 47,000 workers in Q1 2026 alone, according to Layoffs.fyi data. That's a 340% increase from Q1 2025. And 68% of companies cited 'AI transformation' or 'automation initiatives' as the primary reason.

The cuts are concentrated in specific areas:

<ul> <li>Customer service and support (down 41% across major tech firms)</li> <li>Data annotation and content moderation (34% reduction)</li> <li>Junior software development roles (28% fewer openings)</li> <li>Accounting and finance operations (19% cuts)</li> <li>Recruiting and HR coordination (24% reduction)</li> </ul>

Oracle's 10,000 layoffs hit their customer experience division hardest. They've deployed AI agents that can handle 80% of tier-1 support requests. The agents work 24/7, don't take breaks, and cost about $2,000 per month versus $65,000 annually for a human support rep.

Meta's cuts targeted their Reality Labs content moderation teams. Their new AI models can flag problematic content with 94% accuracy (they say). That's three percentage points better than their human teams averaged. So those 3,600 people? Gone.

Block eliminated their fraud analysis team. Their machine learning models now catch suspicious transactions in real-time, something their analysts used to do manually. The AI reviews 100,000 transactions per second. A human team of 1,000 couldn't touch that speed.

<h2>Who's Going All-In on AI</h2>

It's not just these three. I've been tracking the AI adoption patterns across tech companies, and the aggressive movers are:

Salesforce went from 73,000 employees in 2024 to 61,000 today. Their Einstein GPT platform handles what used to require armies of implementation consultants. They're not hiding it either. CEO Marc Benioff said in January that AI will help them "do more with less."

Google eliminated 12,000 positions in 2025 and another 8,000 in early 2026. Their Gemini models now write significant portions of code, manage ad campaigns, and even generate YouTube video descriptions. Junior roles vanished first.

IBM announced they're pausing hiring for 7,800 roles they believe AI can handle within 5 years. Not cutting yet, just not refilling positions as people leave. That's almost scarier because it's calculated.

SAP cut 8,000 jobs last year while simultaneously increasing their AI engineering headcount by 2,400. They're not getting smaller, they're reshaping. Different skills, different roles, different people.

And here's the kicker: these companies are all hiring. Oracle posted 1,200 new positions last month. Meta's got 800 openings. But they're all AI engineers, machine learning specialists, and prompt engineering leads. The median salary? $185,000. The catch? You need skills most laid-off workers don't have.

<h2>The Jobs Getting Hammered</h2>

Let's be specific about what's dying fast:

Data entry and processing: AI can ingest documents, extract information, and populate databases at superhuman speed. Companies like UiPath and Automation Anywhere are selling these capabilities as off-the-shelf products. If your job is moving information from one system to another, you've got months, not years.

Junior developer roles: GitHub Copilot, Cursor, and Replit's AI agent can generate functional code from plain English descriptions. Senior devs are now 3-5x more productive. Companies don't need as many junior developers to maintain codebases anymore. Oracle specifically cut 2,000 junior engineering positions.

Content moderation: Meta's laying off thousands of moderators because their AI models can scan images, video, and text at scale. The AI doesn't get PTSD from reviewing horrible content either. It's cheaper and (they argue) more consistent.

Customer support tiers 1-2: Chatbots used to be terrible. Now they're actually good. Oracle's AI agents resolve 80% of customer inquiries without human intervention. The remaining 20% gets escalated to specialized experts. That middle layer of support staff? Being eliminated.

Translation and localization: Neural machine translation is now good enough for most business content. Companies are cutting translation teams and just running content through DeepL or Google Translate with a quick human review. Meta eliminated 400 localization positions.

Basic accounting tasks: AI can now reconcile accounts, process invoices, and flag anomalies. Oracle cut 800 finance operations roles. Block eliminated their entire accounts payable team and replaced them with automated workflows.

Notice a pattern? These are all tasks that are: <ul> <li>Repetitive and rule-based</li> <li>Don't require deep business context</li> <li>Have clear right/wrong answers</li> <li>Can be measured objectively</li> </ul>

If your job fits that description, you're in the danger zone.

<h2>Where the New Jobs Actually Are</h2>

The good news (yes, there's some): AI is creating new roles. They're just wildly different from what's being eliminated.

AI Trainers and Validators: Someone needs to teach these AI models what good looks like. Meta is hiring 200 "AI behavior specialists" to review model outputs and provide feedback. Pay range: $85,000-$120,000. Required skills: domain expertise in content moderation, no coding needed.

Prompt Engineers: Writing effective prompts is becoming a real skill. Oracle's hiring 150 prompt engineers to improve their AI agents. Salary: $120,000-$160,000. You don't need a CS degree, but you need to understand how large language models think.

AI Integration Specialists: Companies need people who can take AI tools and actually implement them in business processes. Salesforce is hiring 300 of these roles. It's part technical, part change management. Pay: $95,000-$140,000.

Synthetic Data Creators: AI models need training data. Creating high-quality synthetic datasets is now a job. IBM posted 80 positions for synthetic data engineers last month. Salary: $110,000-$150,000.

AI Ethics and Safety Roles: Someone needs to make sure these AI systems aren't making terrible decisions. Every major tech company is building AI safety teams. These roles pay $130,000-$180,000 and require both technical knowledge and policy experience.

But here's the brutal truth: there are maybe 5,000 of these new jobs versus 47,000 layoffs this quarter. The math doesn't work out. Not everyone lands on their feet.

<h2>What You Need to Do Right Now</h2>

Stop waiting to see what happens. That's the worst possible strategy.

First: Take an honest look at your role. Could AI do 80% of what you do today? If you're not sure, that's probably a yes. We built an AI Career Risk Assessment that tells you exactly where you stand and what specific skills you need to develop. Takes 10 minutes, gives you a concrete risk score.

Second: Learn to work WITH AI, not against it. The people keeping their jobs aren't the ones ignoring AI. They're the ones who figured out how to be 3x more productive using AI tools. If you're in customer support, learn how to manage AI agents. If you're in development, master GitHub Copilot. If you're in marketing, get fluent with AI content tools.

Third: Build skills AI can't easily replicate (yet). That means: <ul> <li>Strategic thinking and business judgment</li> <li>Complex stakeholder management</li> <li>Creative problem-solving for novel situations</li> <li>Emotional intelligence and relationship building</li> <li>Cross-functional leadership and influence</li> </ul>

Oracle didn't lay off their senior solution architects. They cut the junior people doing repetitive implementation work. Meta kept their product strategists and community managers who understand cultural nuance. They eliminated the routine content reviewers.

Fourth: Get specific technical skills fast. You don't need a four-year degree. You need practical capabilities employers want right now: <ul> <li>How to write effective prompts and evaluate AI outputs</li> <li>Basic Python for data analysis and automation</li> <li>Understanding of how AI models work (you don't need to build them, just understand their capabilities and limits)</li> <li>Experience with specific AI tools in your industry</li> </ul>

Coursera, Udacity, and LinkedIn Learning have courses you can finish in weeks, not years. Focus on applied skills, not theory.

Fifth: Network like your career depends on it (because it does). The next wave of layoffs is coming. When it hits, having strong relationships with hiring managers, recruiters, and other professionals in your field makes all the difference. Don't wait until you're laid off to start building those connections.

Sixth: Build a financial cushion. If you're in a high-risk role, assume you've got 12-18 months before your position is vulnerable. Save aggressively. Cut unnecessary expenses. Have a plan B.

<h2>The Uncomfortable Reality</h2>

Tech companies are going to keep doing this. The economics are too compelling. Why pay $65,000 plus benefits when an AI agent costs $2,000 per month with no benefits, no vacation, and no sick days?

And it's not stopping at tech companies. Every industry is watching these results. Healthcare, finance, retail, manufacturing. They're all taking notes. The CEO of a major retail chain told me last month they're planning to eliminate 40% of their corporate roles by 2028 through AI automation. They're just waiting to see how the tech layoffs play out first.

This isn't a temporary efficiency drive. It's a permanent restructuring of how work gets done.

The people who survive this transition won't be the ones with the most experience in traditional roles. They'll be the ones who adapted fastest, learned new skills aggressively, and positioned themselves as AI-enhanced workers rather than AI-replaceable workers.

Take the assessment. Figure out your risk level. Build your skills. And do it now, not after the next round of layoffs hits your company.

Because Meta, Oracle, and Block just showed everyone else the playbook. And a lot of companies are going to copy it.

14,000+ Tech Workers Cut as Oracle, Meta, and Block Bet Big on AI

Stay Ahead of AI Job Trends