Microsoft is offering buyouts to approximately 60,000 employees — more than a quarter of its 223,000-person global workforce — in what amounts to the largest voluntary separation program in the company's history. The offer spans major divisions including Xbox, Windows, and Copilot, while shielding cybersecurity, datacenter, and marketing teams from eligibility. It is the second major workforce action this year, following an April offer of voluntary retirement to about 7% of its US workforce. The framing from Microsoft's chief people officer Amy Coleman is revealing in its corporate euphemism: giving employees "who may be considering a future outside the company a new choice." Translated: Microsoft wants to shrink headcount in divisions it considers less essential to its AI pivot, and would rather pay people to leave than fire them outright. The excluded divisions — cybersecurity, datacenters — tell you exactly where Microsoft sees its future revenue. Microsoft is not alone. Meta announced plans to cut 10% of personnel (roughly 8,000 jobs) in April. Oracle shed nearly 13% of its workforce — 21,000 positions — in June. Amazon has made repeated cuts, most recently in retail. Across Big Tech, AI investment is being funded in part by reducing the humans who built the previous generation of products. The critical question is whether any of this is working. The evidence so far is thin. A National Bureau of Economic Research working paper surveying 9,000 senior executives worldwide found that 90% reported AI had not yet produced greater productivity at their companies. Meta has already scaled back some layoff plans. Microsoft's own Ryan Roslansky publicly complained about "AI slop" in the workforce — a striking admission from leadership that AI tools are generating noise, not signal. What's emerging is a pattern where tech companies are cutting labor costs on the promise of AI-driven efficiency, banking the savings now, and deferring the productivity proof to later. The buyout structure is gentler than layoffs, but the economic logic is identical: reduce headcount, redirect capital to AI infrastructure, and hope the bet pays off before the organizational damage compounds. The 60,000 figure deserves scrutiny. Not all eligible employees will accept — buyout uptake typically runs 10-30% depending on terms. But the scope signals that Microsoft is willing to lose a substantial fraction of institutional knowledge across its flagship products. Windows, Xbox, and Copilot are not legacy backwaters; they are current revenue generators. Thinning these teams before AI tools can demonstrably replace human judgment is a gamble with real downside risk. The 20-year trajectory here matters. If AI productivity gains arrive within 3-5 years, these cuts look prescient. If the NBER finding holds — that AI remains a tool that augments rather than replaces most knowledge work — then Big Tech will have hollowed out experienced teams, concentrated wealth among shareholders and executives, and externalized the adjustment costs onto workers who took buyouts into a labor market that's simultaneously absorbing cuts from every major tech employer.
