Q2 2026 Tech Layoffs: The AI Replacement Wave Hits Different This Time

The second quarter numbers are in, and they're confirming what many of us feared but hoped wouldn't happen this fast.

43,000 tech workers lost their jobs between April and June 2026. But here's what makes this different from previous rounds: companies aren't hiding behind euphemisms anymore. They're explicitly stating that AI systems can now handle the work.

What's Actually Happening

Salesforce laid off 3,200 people in May, with CEO Marc Benioff stating their new Agentforce AI can handle customer service queries that previously required 8-12 human agents. The company's earnings call mentioned "AI efficiency gains" 47 times.

IBM cut 2,800 positions in April. Not restructuring. Not rightsizing. The internal memo literally said: "Our AI systems have reached capability parity with human analysts for these functions."

And it's not just the usual suspects. Intuit (1,800 jobs), SAP (2,400 jobs), and Cisco (4,100 jobs) all announced similar moves within a three-week period in May.

We're seeing a pattern we haven't seen before: coordinated replacement, not reduction.

The Numbers Tell A Story

Here's what we're tracking across Q2:, Customer service/support roles: 12,400 positions eliminated (28.8% of total), Software QA and testing: 8,900 positions (20.7%), Data entry and analysis: 6,200 positions (14.4%), Content moderation: 4,800 positions (11.2%), Junior/mid-level developers: 3,700 positions (8.6%), Technical writing and documentation: 2,900 positions (6.7%), HR screening and recruiting: 2,100 positions (4.9%), Other technical roles: 2,000 positions (4.7%)

The average severance package decreased too. Q1 2026 averaged 16 weeks. Q2 dropped to 12 weeks. Some companies are citing "market adjustment" (which is corporate-speak for "everyone else is doing it, so we can too").

Who's Moving Fastest

Google's been the most aggressive. They've reduced their customer operations team by 40% since January, replacing them with Gemini-powered support systems. Their own data shows customer satisfaction scores dropped only 2.3 points, which apparently is acceptable.

Microsoft announced their "AI Co-pilot Workforce Transition" in early May. Translation: 5,200 people across Azure support, documentation, and testing teams. They're retraining 1,800 of them. The other 3,400 got severance and LinkedIn Premium for a year.

Meta's been quieter but just as aggressive. They've eliminated most of their content moderation team in Southeast Asia (2,100 people), replacing them with updated AI models that Mark Zuckerberg claims are "95% as accurate and 24/7 available."

Amazon Web Services cut 2,600 technical support roles. Andy Jassy said on the earnings call that their AI assistant now resolves 73% of customer issues without human intervention, up from 31% in Q4 2025.

The Sectors Getting Hit

Technology/Software: Obviously leading with 18,900 jobs (44% of total Q2 layoffs). But what's interesting is the breakdown: established enterprise companies are cutting faster than startups right now.

Financial Services: 8,200 positions, mostly in fraud detection, loan processing, and customer service. JPMorgan Chase alone cut 1,400 roles, explicitly stating their AI models now handle 82% of routine customer inquiries.

Healthcare Tech: 6,100 jobs. Medical coding, billing, transcription services getting automated rapidly. Nuance (owned by Microsoft) eliminated 900 medical transcription roles in April.

E-commerce/Retail Tech: 4,900 positions. Shopify cut 800 people from their merchant support team. Their AI support bot now handles the work.

Telecommunications: 2,800 jobs, mostly customer service and network monitoring roles.

Media/Publishing: 1,600 positions. Content moderation and basic copywriting roles disappearing fast.

Education Tech: 500 positions. Smaller numbers but growing fast, mostly tutoring and grading automation.

Real Voices From Real People

I've been talking to displaced workers all quarter. Here's what they're telling me:

"I trained the AI that replaced me." That's from Marcus, who spent 6 years at a customer service software company. He helped label training data for their new AI agent. Three months later, his entire team was gone.

Jenna worked in QA testing for 8 years. "The AI finds bugs faster than I could. That's just true. But it doesn't understand why users will be frustrated by certain UX choices. Nobody seems to care about that anymore."

Raj was a technical writer. "They kept my manager and one senior writer. Everyone else got cut. They're using Claude and GPT-4 to generate documentation now, and the senior writer just edits it. One person doing what seven of us did."

The pattern: companies are keeping 1-2 senior people per team to manage AI systems and cutting everyone else.

But Here's What Companies Aren't Telling You

The quality is dropping. We're seeing it in customer complaints, bug reports, and user satisfaction scores. Just not enough for companies to reverse course yet.

Salesforce's customer sat scores are down 8% since they cut their support team. But their costs are down 34%. From a CFO perspective, that's still a win.

Google's AI support can't handle complex, multi-step issues. So those get escalated, and now their remaining human agents are drowning. Average resolution time has doubled. But total support costs are down, so leadership calls it success.

The technical debt is piling up. When you cut QA teams and rely on AI testing, you catch the obvious bugs but miss the subtle ones. Three major software companies had significant outages in Q2 that engineers privately blamed on reduced human oversight.

The Jobs That Are Growing

Okay, not all doom. Some roles are actually expanding:

**AI Trainers and Evaluators**: Companies need people to continuously improve their AI systems. But here's the catch: these jobs pay 30-40% less than the roles they're replacing and are often contract positions, not full-time.

**AI Ethics and Compliance Officers**: Growing fast, especially in healthcare and finance. Companies are getting nervous about AI mistakes and regulatory exposure. These are good jobs, but there are maybe 200 openings nationwide versus 43,000 layoffs.

**Prompt Engineers**: Yes, still growing. The job market data shows 3,400 open positions in Q2, up 18% from Q1. Average salary is $145,000. But the skill requirements are getting more specialized. You need to understand model architectures, not just write good prompts.

**AI Integration Specialists**: Companies need people who can implement and customize AI systems for their specific workflows. About 2,800 open positions. These require both technical skills and deep domain knowledge.

**Data Quality Analysts**: Companies are realizing that AI is only as good as the data it trains on. Growing field, roughly 1,900 openings in Q2.

But let's be honest about the math: maybe 8,000-10,000 new AI-related jobs emerging versus 43,000 traditional tech jobs eliminated. That's not replacement, that's displacement.

What This Means For Your Career

If you're in customer service, QA, technical writing, data entry, or content moderation, you need to be planning your next move now. Not in six months. Now.

The companies moving to AI aren't going back. This isn't a temporary efficiency play. They're redesigning entire workflows around AI-first operations.

Here's what I'm telling people:

**Assess your actual risk**: Take our AI Career Impact Assessment. It'll give you a specific readiness score for your role and industry. Don't guess. Get data.

**Document everything you know**: Your domain expertise is valuable. Start creating a portfolio of your specialized knowledge. The AI knows the general patterns; you know the exceptions and edge cases.

**Learn to work alongside AI**: The people keeping their jobs aren't the ones who refuse to touch AI. They're the ones who became experts at managing and improving AI systems. Get hands-on now.

**Build relationships outside your company**: Your network is your safety net. People find jobs through people, especially when AI can screen resumes better than humans now.

**Consider parallel skills**: If you're in QA, maybe learn security testing or accessibility testing (both harder to automate). If you're in customer service, maybe move toward customer success or account management (relationship-heavy roles).

**Get financially prepared**: If you haven't already, build that emergency fund. Three months of expenses minimum. Six is better. The job market for displaced workers is brutal right now.

The Q3 Outlook

Based on what we're hearing from industry contacts and leaked company memos:

Expect another 35,000-40,000 tech layoffs in Q3. Oracle's planning a major restructuring in July. Accenture's got something coming in August (they're calling it a "strategic workforce alignment").

The automation is moving up the skill chain. We're starting to see junior and mid-level developer roles getting cut. Not yet senior roles, but give it another quarter or two.

Companies are getting bolder about the messaging. They're not hiding behind "reorganization" language anymore. They're straight up saying: "AI can do this job."

And honestly? Sometimes they're right. Sometimes the AI can do the job at 85% quality for 5% of the cost. That's good enough for most companies.

What You Should Do This Week

Don't wait to see if your company is next. They probably are.

Take the AI Career Impact Assessment at aicrisis.org. It's free, takes 10 minutes, and gives you a personalized risk score plus specific recommendations.

Update your resume and LinkedIn. Highlight projects where you worked with AI tools, improved processes, or solved unusual problems (the stuff AI can't do yet).

Reach out to three people in your network. Just check in. Ask about their companies. Be genuine, not transactional. But rebuild those connections.

Start a learning project. Pick one AI tool relevant to your field and become the expert. Document what you learn. That becomes your proof that you can adapt.

Look at your finances. For real. What's your runway if you lost your job tomorrow? What could you cut? What would you need to figure out?

The hard truth: Q2 2026 was probably not the peak. This is the beginning of a multi-year transition. The workers who survive are the ones who started preparing yesterday.

And if you're already displaced? You're not alone. 43,000 other people are figuring this out too. The jobs that are coming next will be different from the ones we lost. But they will come. The question is whether you'll be ready for them.