In April and May 2026, two of the world’s most valuable technology companies made the same bet: fewer people, more machines. Meta Platforms confirmed it is eliminating roughly 8,000 positions, and Microsoft began offering voluntary buyouts to approximately 8,750 U.S. employees. Both companies are simultaneously pouring record sums into artificial intelligence infrastructure, a combination that marks one of the sharpest workforce pivots in Silicon Valley history.
Meta: cutting 8,000 while doubling down on AI
Meta disclosed the layoffs alongside its fourth-quarter and full-year 2025 earnings, published through its official newsroom. In an internal memo to staff, CEO Mark Zuckerberg wrote: “I’ve decided to raise the bar on performance management and move out people who aren’t meeting expectations over the course of a year.” Portions of the memo, reported by Bloomberg and other outlets, framed the reductions not as belt-tightening but as a deliberate reallocation: roles that do not feed Meta’s AI ambitions are being phased out so the company can fund the ones that do.
The spending figures reinforce that framing. According to Meta’s fourth-quarter 2025 earnings release, available through its investor relations page, the company guided capital expenditures of $60 billion to $65 billion for 2025, the vast majority earmarked for data centers, GPU clusters, and AI model training. That figure alone dwarfs whatever salary savings 8,000 layoffs would generate, which suggests the cuts are driven as much by organizational focus as by cost reduction.
Meta ended 2024 with approximately 72,000 employees, according to its annual filings. At that headcount, 8,000 eliminations would represent roughly 11% of the workforce rather than the 10% figure initially circulated in some reports. The distinction matters for gauging the depth of the restructuring.
What Meta has not disclosed is where the deepest losses will land. The company has not published severance terms, departure timelines, or a breakdown of how many AI-specific hires it plans to add. Employees in functions like trust and safety, marketing, and business operations are left reading tea leaves. Internal transfers to AI teams remain a possibility for some workers, but Meta has not said how many roles will be filled internally versus through outside recruiting.
This is not Meta’s first mass reduction. The company cut 11,000 jobs in November 2022 and another 10,000 in early 2023, both during a broader pullback Zuckerberg labeled a “year of efficiency.” Those rounds trimmed a workforce that had ballooned during the pandemic hiring surge. The 2026 cuts carry a different tone: they are happening while the company is actively spending more, not less, a signal that the goal is transformation rather than contraction.
Microsoft: buyouts instead of pink slips
Microsoft is taking a softer approach, at least on the surface. The company plans to offer voluntary separation packages to about 8,750 U.S. employees, according to reporting by the Associated Press, which cited people familiar with the plan. The AP report characterized the figure as roughly 7% of Microsoft’s domestic workforce, though Microsoft does not routinely disclose its U.S.-specific headcount, so that percentage rests on an estimated base of approximately 125,000 domestic employees. A voluntary buyout gives workers a financial cushion and a choice, but the sheer scale of the offer signals that leadership views significant portions of the current workforce as misaligned with the company’s strategic direction.
That direction is hard to miss. During its fiscal-year 2025 earnings calls, Microsoft indicated it expected approximately $80 billion in capital spending for the fiscal year ending June 2025, much of it flowing into Azure data centers optimized for AI workloads and into its multibillion-dollar partnership with OpenAI. The company has woven generative AI into nearly every major product line: Copilot features in Office and Windows, AI-powered coding tools in GitHub, and AI assistants embedded across Dynamics and Security. Roles tied to legacy product maintenance or functions that AI tools can partially automate appear to be the likeliest candidates for buyout offers.
Key details remain unconfirmed. Because the buyout plan has not been announced through an SEC filing or official press release, the exact terms, acceptance deadline, and targeted business units are still unclear. If too few employees volunteer, Microsoft could face a difficult choice: accept a slower restructuring or pivot to involuntary layoffs, as it did in January 2023 when it cut 10,000 positions. The company has not commented publicly on contingency plans.
Where the money is actually going
Between them, Meta and Microsoft are on track to spend north of $140 billion on capital expenditures in a single fiscal year. The bulk of that goes to physical infrastructure: land, buildings, power systems, networking equipment, and the specialized chips that train and run large AI models. Those investments create construction and engineering jobs, but they employ far fewer people per dollar than traditional software teams do. A $1 billion data center might require a few hundred workers to operate once built; the same amount spent on salaries could support thousands of engineers for a year.
That math helps explain why the layoff and hiring numbers will not cancel each other out. Even if both companies add thousands of AI-focused roles, the net effect on total headcount is likely negative in the near term. For workers whose skills sit outside machine learning, data engineering, or AI product development, the implication is blunt: the fastest-growing budget lines in Big Tech are not designed to employ them.
“We’re seeing a structural decoupling of revenue growth from headcount growth,” said Gil Luria, a senior research analyst at D.A. Davidson, in an April 2026 interview with CNBC. “These companies are telling you with their checkbooks that the next trillion dollars of value will be created by AI systems, not by adding more people.” That framing resonates with policymakers who have started paying attention. In recent months, lawmakers on both sides of the aisle have raised questions about the pace of AI-driven workforce displacement, though no major legislation specifically addressing AI layoffs has advanced through Congress. For the roughly 17,000 workers across Meta and Microsoft now facing job loss or buyout decisions, the political conversation remains well behind the corporate one.
Signals to watch as restructuring plans take shape
For employees at either company, the most important near-term signal is divisional detail. Once Meta identifies which teams are losing headcount and Microsoft reveals which units are targeted for buyouts, the true scope of the restructuring will come into focus. Workers in AI-adjacent roles, such as data annotation, cloud operations, and applied research, may find themselves in higher demand internally even as colleagues in other functions are shown the door.
For investors, the central question is whether AI spending translates into revenue growth fast enough to justify the cost. Meta’s core advertising business remains highly profitable, and AI-driven recommendation improvements have already lifted engagement metrics across Facebook and Instagram. Microsoft’s Azure cloud division has posted accelerating growth tied to AI workloads. But capital expenditures at this scale compress margins in the short run, and Wall Street will scrutinize quarterly results for evidence that returns are materializing on schedule.
What makes this moment distinct from previous tech layoff cycles is the simultaneity of cutting and spending. Two companies that collectively employ more than 300,000 people have concluded, almost in lockstep, that their futures depend on AI to a degree that justifies shrinking their existing workforces while writing the largest infrastructure checks in their histories. That is not a temporary cost-cutting phase. It is a structural redefinition of what these firms consider essential work, and it is unfolding faster than most employees, regulators, or competitors anticipated.
