In the same week that Big Tech announced roughly $650 billion in combined AI capital expenditure for 2026, two of the industry’s biggest employers quietly showed 20,000 people the door.
Meta confirmed 8,000 layoffs starting May 20 — the largest single round in the company’s history — while simultaneously freezing 6,000 open positions. Microsoft, not to be outdone in grim corporate milestones, offered voluntary buyouts to 8,750 U.S. workers. It’s the first time in Microsoft’s 51-year history they’ve done that.
The message couldn’t be clearer if it was written in GPU compute cycles: the money for AI has to come from somewhere, and that somewhere is people.
The Numbers Are Staggering
Let’s stack them up:
- Meta: 8,000 layoffs + 6,000 frozen roles = 14,000 positions gone. AI capex for 2026: $115–135 billion.
- Microsoft: ~8,750 voluntary buyouts (7% of U.S. workforce). AI capex for 2026: ~$80 billion.
- Combined 2026 tech layoffs so far: 92,000+ and counting — nearly double all of 2025.
- AI attribution: Analysts estimate 30–50% of 2026 tech layoffs cite AI as a primary factor, up from 8% in 2024.
And then there’s the spending side. Meta, Microsoft, Amazon, Google, and others are collectively projecting around $650 billion in AI infrastructure investment this year. That’s a 67% increase from 2025. Someone’s paying for those data centres, and apparently it’s the people who used to work at them.
”Rightsizing” or Wrong Direction?
Executives frame these cuts as “rightsizing” after pandemic over-hiring. And sure, some of that is real — tech did bulk up irresponsibly during COVID. But the timing is suspicious, isn’t it? Meta’s May 20 layoffs come with explicit language about reallocating resources toward AI. Microsoft’s unprecedented buyout program coincides with the company writing 30% of its code with AI in some projects.
Call it what you want. The pattern is unmistakable: cut headcount, pour the savings into AI infrastructure, repeat.
CTOs are already telling anyone who’ll listen. One tech executive posted about plans to cut a 180-person engineering team to 47 — replacing mid-level roles with AI agents and offshore workers at a third of the cost. Entry-level and mid-level positions? Gone. Seniors managing AI outputs? That’s the new org chart.
Who Gets Hit Hardest
This isn’t hitting equally. The Dallas Fed confirmed what everyone suspected: entry-level roles are shrinking while experienced worker wages are rising. It’s a brutal double standard — we’ll pay senior engineers more because they can wrangle AI systems, but we won’t invest in training the juniors who could become them.
The job displacement numbers from MIT are equally grim: $2 trillion in projected wage displacement, with coding-first roles in the crosshairs. And Dell’s 11,000 engineer layoffs happened the same quarter their AI revenue hit record highs.
If you’re early career, the message is bleak. If you’re mid-career, it’s terrifying. If you’re a senior who can work with AI tools, congratulations — you’re the last human they’ll keep.
The NZ Angle
No major NZ-specific layoffs have been reported yet, but the pressure ripples through. Multinationals operating in Aotearoa — Microsoft, Meta, Amazon — are cutting globally, and NZ staff aren’t exempt. More importantly, the signal matters: if the biggest companies on Earth are replacing humans with AI at this pace, NZ’s smaller tech ecosystem feels it faster.
NZ’s tech sector relies heavily on overseas contracts and platforms. When those platforms automate, the local jobs that serviced them disappear. And our regulation? Still playing catch-up. NZ’s AI compliance framework is evolving but hasn’t caught up with the velocity of what’s happening.
The kiwi engineer posting on X about AI outpacing juniors? That’s not a hypothetical. That’s the view from Wellington.
The Honesty Problem
Here’s what really grates: no company will just admit they’re cutting jobs for AI. Every press release talks about “efficiency,” “restructuring,” “focusing on priorities.” Microsoft frames buyouts as voluntary — as if 8,750 people simultaneously decided they’d rather not work at Microsoft.
The transparency gap is staggering. Some CEOs are starting to be more honest — Block’s Jack Dorsey explicitly cited AI when cutting 40% of staff — but most still hide behind corporate euphemisms.
Meanwhile, half of American workers already fear AI will take their jobs, and companies are proving those fears right faster than anyone predicted.
🔍 The Bottom Line
The AI labor crisis isn’t coming. It’s here. 92,000+ tech jobs gone in early 2026, $650 billion flowing into AI infrastructure, and the companies doing both insist there’s no connection.
There is. We all see it. The only question is whether we’ll have honest policy responses — retraining programs, transition support, AI displacement taxes — or whether we’ll keep pretending the free market will sort it out while millions of careers disappear.
For workers navigating this: upskill in AI tools now, not later. Not because it’s fair — it isn’t — but because the alternative is waiting for a system that’s not coming to save you.
