The $700 Billion Bluff: Why Companies Are Cutting Workers to Fund AI That Isn't Delivering

2026-06-22 · Nia

Cloudflare posted record quarterly revenue of $639.8 million, then cut 1,100 workers the same day. Meta reported $26.8 billion in net income and eliminated 8,000 roles. Block projected $12 billion in gross profit while slashing 40% of its workforce.

These aren't struggling companies making hard choices. They're the most profitable enterprises on the planet, cutting workers at peak performance to fund AI infrastructure that hasn't proven it can replace the people being fired.

And the data says it's not working.

The Numbers Don't Lie

More than 142,000 tech jobs were eliminated in the first five months of 2026. AI is now the most-cited reason for corporate layoffs, with an estimated 87,714 jobs attributed to AI so far this year — up from 54,836 for all of 2025.

Meanwhile, hyperscalers like Amazon, Microsoft, Alphabet, and Meta have committed roughly $700 billion to AI infrastructure in 2026 alone. Nearly double what they spent last year.

The money has to come from somewhere. Right now, it's coming from payroll.

The Gartner Bombshell

Here's where the story collapses. A May 2026 Gartner study surveyed 350 global executives at companies with at least $1 billion in annual revenue. Among organizations piloting or deploying AI, roughly 80% had reduced their workforce.

The critical finding: none of it correlated with better returns.

Companies that cut workers got the same financial outcomes as companies that didn't. Zero difference. The layoffs are theater — budget room dressed up as strategy.

As Gartner's Helen Poitevin put it, many CEOs implement layoffs to demonstrate quick AI returns, but this approach is "misplaced." Workforce reductions might create budget room, but they don't generate returns.

It gets worse. Gartner predicts that up to 30% of roles displaced by AI will be rehired by 2029, often at higher cost due to rehiring expenses, compensation premiums, and recruitment fees. Companies are paying twice for the same positions — once to fire people, and again to hire them back.

The Companies That Actually Get It

The organizations seeing real AI returns aren't cutting headcount. They're doing something Gartner calls "people amplification" — aggressively investing in skills, new roles, and operating models that empower humans to guide and scale autonomous systems.

The PwC 2026 Global AI Jobs Barometer confirms this with striking data. Companies most exposed to AI are seeing faster headcount growth (52% versus 36% for least AI-exposed companies), higher wage growth (24% versus 17%), and labor productivity gains of 163% for the "super-star" adopters.

Read that again: the companies winning with AI are hiring more people, not fewer.

The barometer also reveals a "two-track" labor market emerging. "Professionalized" roles — where AI automates routine work and humans bring judgment, creativity, and leadership — are growing twice as fast as "democratized" roles where AI lowers skill barriers. The wage premium for AI skills has hit 62%, up from 57% last year is a signal too loud to ignore.

We covered this divide in depth in The Two-Tier Workforce — and it's only accelerating.

Why Smart Leaders Fire. And Why Smarter Ones Don't.

There's a practical reason why layoffs backfire as an AI strategy: AI needs institutional knowledge to function.

It needs people who understand the business, the customers, the context behind decisions. When you fire those people, the AI can't operate properly. Some companies have already had to rehire for roles they previously eliminated after discovering that automation could handle only part of the work.

Look at the layoff list. Cisco cut 4,000 jobs citing AI adoption. Snap cut 1,000 expecting AI to do the work of larger teams. Oracle is reportedly cutting 20,000-30,000 employees. Coinbase reduced teams to single individuals expected to use AI agents for work previously done by many.

But there's a disconnect nobody talks about: only 1% of recently laid-off U.S. workers named AI as the primary reason for losing their jobs. They cite restructuring, cost-cutting, and capital reallocation. AI is the excuse, not the cause.

This is what we flagged in Corporate AI Strategy Theater — CEOs using AI as cover for plain old cost-cutting, dressed up in innovation language.

The Entry-Level Crisis Nobody's Preparing For

PwC's data reveals something particularly troubling. Entry-level roles most exposed to AI are now seven times more likely to require traditionally senior-level skills — leadership, strategic thinking, judgment. Job openings for these "seniorised" entry-level roles have grown by 35% since 2019. Other entry-level roles? Down 10%.

This creates a brutal catch-22: companies fire experienced workers to fund AI, then can't find juniors with the skills to work alongside AI. The talent pipeline is breaking at both ends.

A new report suggests 99% of CEOs expect AI-driven workforce reductions within the next two years, with younger workers particularly vulnerable. But the evidence says the companies that resist this impulse — the ones that retrain instead of replace — will be the ones still standing in 2030.

What This Means for You

If you're a leader: stop thinking of AI as a replacement for headcount. The ROI on AI-driven layoffs is zero. Redesign roles. Invest in training. Build the skills that make AI actually useful.

If you're a worker: your institutional knowledge is more valuable than ever. The AI literacy wage premium is real — workers with AI fluency see triple the job security. Learn the tools, but don't forget that what you know about the business is the part AI can't replicate.

If you're a founder: this is your opportunity. While giants burn through $700 billion on infrastructure and hemorrhage institutional knowledge, lean teams that build intelligently with AI are moving faster with fewer resources.

The $700 billion question isn't whether AI will transform work. It will. The question is whether cutting the humans who make AI useful is the way to get there.

The data says no.