Meta is eliminating approximately 8,000 jobs — about 10% of its global workforce — in the company’s largest restructuring since Mark Zuckerberg’s 2022 “Year of Efficiency” campaign. The cuts take effect May 20, and the company is also scrapping plans to fill roughly 6,000 open roles that had already been approved for hiring.

The stated reason is straightforward: free up capital for AI.

Meta’s 2026 capital expenditure guidance sits at $115 billion to $135 billion — nearly double the $72 billion it spent in 2025. The company has reorganised major product and engineering teams into AI-focused units under newly appointed Chief AI Officer Alexandr Wang, whose Superintelligence Labs is now the centrepiece of Meta’s long-term strategy.

“I think that 2026 is going to be the year that AI starts to dramatically change the way that we work,” Zuckerberg said ahead of the announcement.

Meta’s Chief People Officer Janelle Gale delivered the news in an internal memo, describing the restructuring as an effort to “drive a step change in engineering productivity and product quality” and to offset the cost of new AI investment. Additional layoffs are expected in the second half of 2026.

The Scale of the Bet

Meta’s $115-135 billion capex commitment is not a rumour or a projection — it’s public guidance. The money goes to AI data centres, compute capacity, custom silicon, and the engineering talent to run it all. For context, that figure exceeds the annual GDP of more than 100 countries.

What makes this round of cuts different from 2022 is the directionality. Three years ago, Meta cut headcount because the business had over-hired during the pandemic boom. This time, the cuts are deliberate resource reallocation: fewer people, more machines.

Zuckerberg has made clear that Meta’s competitive advantage in the next phase of tech will be built on AI infrastructure, not human labour at scale. The Superintelligence Labs org — reporting directly to the CEO through Alexandr Wang — is the clearest signal of where the company is placing its long-term bets.

What This Means for Business

AI spending is cannibalising headcount at the top of the market. Meta, Amazon, and others are explicitly trading employees for compute. This isn’t a distant trend — the memos have been sent.

The pressure flows downstream. When the largest tech companies restructure around AI productivity, the implicit message to everyone else is: if you haven’t asked what AI can do in place of headcount, someone with access to capital already has. The businesses that figure this out first keep margin; the ones that don’t find it harder to compete on cost.

Workforce restructuring doesn’t require Meta’s scale. Most small and mid-sized businesses can’t spend $135 billion on AI infrastructure. But the underlying logic — identify where repetitive, processable work is being done by humans, and figure out which of it AI can handle — applies at any size. The difference is that at smaller scale, the risk is lower and the implementation is faster.

The “AI vs. jobs” framing misses the practical question. The real question for most business leaders isn’t philosophical. It’s operational: which tasks in your business take significant time and produce predictable outputs? Those are the ones worth examining.

For a professional services firm, that might be client intake and scheduling. For a retailer, it might be inventory reporting. For a trade business, it might be quote generation and follow-up. None of these require Superintelligence Labs. They require clear workflows, the right tools, and someone who can tie them together.

The Microsoft Footnote

Microsoft is moving in a parallel direction, offering voluntary buyouts to employees across multiple divisions. While Microsoft hasn’t announced companywide cuts at Meta’s scale, the pattern is consistent: mature tech businesses with heavy AI investment programmes are actively managing their headcount down.

Both companies are building the same argument to investors: AI productivity improvements will more than offset the short-term cost of restructuring. The test of that argument plays out over the next 12-24 months.

EDNA’s Take

Meta’s announcement is the clearest signal yet that the trade-off between human labour and AI capability is no longer hypothetical. It is now a line item in a $135 billion capital plan.

For business owners, the response isn’t panic — and it isn’t copying Meta. It’s asking an honest question: what would your business look like if your team spent less time on repetitive work and more time on the things only humans do well?

That’s not a technology question. It’s an operational one. The technology is increasingly available, and increasingly affordable. The gap is usually in knowing where to start.

---

Want the practical version of this? The free Working With Claude field guide covers the full Claude ecosystem, Claude Code, and how to roll it out across a real business. Download it here.