The $135 Billion Squeeze: Inside Meta's Plan to Cut 20% of Its Workforce to Fund an AI Moonshot

Meta Platforms is reportedly considering eliminating up to 20% of its nearly 79,000-person workforce — roughly 15,800 jobs — as CEO Mark Zuckerberg races to redirect the company's resources toward a breathtakingly expensive bet on artificial superintelligence [1][2]. The potential cuts, first reported by Reuters on March 14, would represent the largest single restructuring in Meta's history and signal a dramatic new phase in Big Tech's AI arms race, one where human workers are increasingly viewed not as assets to be retained but as costs to be optimized away.

The Scale of What's Coming

Senior leaders across Meta have been told to begin planning for broad reductions, though no final timeline or headcount number has been set [3]. Meta spokesperson Andy Stone dismissed the reports as "speculative reporting about theoretical approaches," but the scope of the internal discussions — affecting every major division — suggests the planning is well advanced [4].

If Meta proceeds with a 20% cut, the reduction would dwarf even the company's brutal 2022-2023 restructuring, when Zuckerberg eliminated 21,000 positions across two rounds of layoffs during what he called the company's "year of efficiency" [5]. That earlier bloodletting slashed headcount from a pandemic-era peak of 86,482 to 67,317, before Meta began rehiring and grew back to approximately 79,000 by the end of 2025 [6].

The irony is stark: many of the workers now facing the chopping block were hired during the post-layoff recovery, recruited specifically to staff up Meta's AI and infrastructure ambitions. The company is now preparing to cut the very workforce it spent the last two years rebuilding.

Source: MacroTrends / Meta SEC Filings

Data as of Mar 14, 2026CSV

Follow the Money: $135 Billion in Capital Spending

The driving force behind the contemplated layoffs is a capital expenditure plan of almost incomprehensible scale. Meta has guided investors to expect between $115 billion and $135 billion in capital spending for 2026 — nearly double the $72.2 billion it spent in 2025, which itself was a record [7][8].

The money is flowing into two massive data center projects: Prometheus, a 1-gigawatt facility set to begin operations this year, and Hyperion, designed to scale up to 5 gigawatts across multiple phases [9]. Meta has also signed a $100 billion multi-year agreement with AMD to deploy up to six gigawatts of AI computing infrastructure, with first shipments expected in the second half of 2026 [9].

This spending is anchored in Zuckerberg's vision for what he calls "personal superintelligence" — AI systems built not just for enterprise or research applications but for individual users [9]. To lead this effort, he hired Scale AI founder Alexandr Wang as Meta's first chief AI officer in a deal that included a $14.3 billion acquisition of Wang's startup [10]. The company has consolidated its AI efforts into a new division called Meta Superintelligence Labs (MSL), signaling that the pursuit of artificial general intelligence is now the company's central strategic priority [9].

The Metaverse's $80 Billion Funeral

The pivot to AI has come at the direct expense of Meta's once-defining bet on the metaverse. Reality Labs, the division that inspired the company's very name change from Facebook in 2021, has now accumulated close to $80 billion in operating losses since its creation [11]. In Q4 2025 alone, the division recorded $6 billion in losses on just $470 million in revenue [11].

Meta has responded by slashing the metaverse budget by 30%, laying off approximately 1,500 Reality Labs employees (about 10% of the division), shutting down VR game studios, and freezing or canceling advanced headset development [12][13]. The company is redirecting remaining hardware teams toward AI-powered wearables — specifically, the Ray-Ban Meta AI glasses, which have sold more than two million units and represent the only consumer hardware success to emerge from the metaverse era [13].

The shift has sparked fears of a "VR winter" across the broader industry, as developers and studios that had bet on Meta's virtual reality ecosystem now face an uncertain future [14].

A Pattern Across the Tech Industry

Meta's contemplated cuts are part of a much larger wave sweeping through the technology sector. According to an analysis by RationalFX, 45,363 tech workers have been laid off globally since the start of 2026, with roughly 9,200 — about 20% — directly attributed to AI implementation and organizational restructuring [15]. If the current pace continues, total tech layoffs could reach 264,730 by year-end, surpassing 2025's total of 245,000 [15].

Amazon has announced 16,000 cuts so far this year. Block eliminated 4,000 positions. eBay cut 800. Pinterest shed 675 [16]. In a survey of 1,000 U.S. hiring managers by Resume.org, 55% said they expect layoffs at their companies in 2026, and 44% identified AI as a top driver [16].

Source: Bureau of Labor Statistics (CES5000000001)

Data as of Mar 14, 2026CSV

The pattern is unmistakable: companies are simultaneously spending record sums on AI infrastructure while cutting the human workers whose roles they believe AI can augment or replace. As Crowdbyte previously reported, ServiceNow CEO Bill McDermott warned that AI agents could push unemployment among recent college graduates into the "mid-30s" within the next couple of years, even as his own company automates 90% of its customer service operations.

The Financial Paradox

What makes Meta's situation particularly striking is that the company is not in financial distress. Far from it. Meta reported $200.97 billion in revenue for full-year 2025, up 22% year-over-year, with Q4 revenue of $59.89 billion beating analyst expectations [17]. Earnings per share of $8.88 in Q4 also exceeded consensus estimates [17]. The advertising business — which still accounts for more than 97% of revenue — continues to benefit from AI-driven improvements to ad targeting and user engagement across Facebook, Instagram, WhatsApp, and Threads [18].

Yet the company's costs are rising just as fast. Total expenses reached $117.69 billion in 2025, up 24% year-over-year, and the 2026 capex guidance of up to $135 billion means Meta will spend more on infrastructure this year than the entire GDP of countries like Morocco or Ecuador [7]. The company generated $43.59 billion in free cash flow in 2025, but its planned capital spending for 2026 could consume roughly three times that figure [8].

Wall Street has so far given Zuckerberg the benefit of the doubt. Meta shares rose as much as 10% after the Q4 earnings report, and the consensus recommendation from 67 analysts covering the stock remains "buy," with a median price target of $840 [19]. But the stock was trading at $613.71 as of March 13, down 3.83%, reflecting growing unease about whether the AI spending binge will actually pay off [19].

Inside Meta: A "Crisis of Trust"

The human toll of Meta's strategic whiplash is playing out inside the company. Employees who survived the 2022-2023 layoffs, only to be told they were the "backbone" of Meta's future, are now watching as the company prepares another massive restructuring [20]. The atmosphere has been described as a "crisis of trust," particularly among high-performing employees who question whether previous rounds of layoffs were genuinely performance-based [20].

In a notable shift, Meta has announced it will scrap its controversial 5% performance-based layoff strategy for 2026, opting instead for departmental reductions tied to strategic priorities rather than individual performance metrics [20]. The move is intended to stabilize morale, but critics inside and outside the company argue that the distinction is cold comfort to workers whose entire departments may be deemed expendable.

The broader cultural impact extends beyond Meta's walls. The tech industry, which for two decades positioned itself as a meritocratic haven offering stability, lavish perks, and mission-driven work, is rapidly becoming a sector where even top performers face existential uncertainty every 12 to 18 months.

The Superintelligence Gamble

The ultimate question hanging over Meta's restructuring is whether Zuckerberg's bet on superintelligence will prove more prescient than his bet on the metaverse — or whether it will produce another $80 billion crater on the company's balance sheet.

The early signals are mixed. Meta's Llama large language model family has positioned the company as a credible AI player, but the rollout has been bumpy. The Llama 4 release in April 2025 was widely seen as disappointing, and the company's largest model, Llama 4 Behemoth, has been repeatedly delayed [9]. Meanwhile, competitors like OpenAI, Google DeepMind, and Anthropic continue to push the frontier, making it far from certain that even $135 billion in spending will be enough to achieve — or maintain — a leadership position.

Zuckerberg has signaled that Meta may not open-source all of its superintelligence models, a potential departure from the open approach that made Llama popular with developers and researchers [21]. That decision, if confirmed, would represent a philosophical reversal and could alienate the developer community that has been one of Meta's strongest AI assets.

What Comes Next

Meta's contemplated layoffs represent more than a cost-cutting exercise. They are the clearest signal yet that the largest technology companies view AI not merely as a product category or a feature enhancement, but as a force so transformative that it justifies restructuring entire organizations around it — including, paradoxically, eliminating many of the workers who were hired to build it.

For the roughly 15,800 Meta employees who may lose their jobs, the message is unambiguous: in the age of AI, no one is safe. Not the metaverse engineers whose division inspired a corporate rebirth, not the content moderators whose work keeps the platforms running, and not the mid-level managers whose coordination roles are precisely the kind that AI is designed to replace.

The tech industry is entering a phase where companies post record revenues, record capital spending, and record layoffs simultaneously. Meta's $135 billion gamble on superintelligence — funded in part by cutting the workers who built the company into a $1.5 trillion enterprise — is the starkest illustration yet of what that future looks like.