Microsoft has offered voluntary severance to about 60,000 employees, roughly a quarter of its workforce, while keeping teams tied to artificial intelligence out of the program. Chief People Officer Amy Coleman described the packages as another way to adjust staffing to changing needs. The move matters because headcount discipline now runs in parallel with very large capital spending on AI infrastructure.
Which roles are covered by the buyout offer
The program applies to employees in the United States, Canada, India and the United Kingdom at vice president level and below. Xbox, Windows, Copilot and Finance are fully included, which puts both product and corporate functions inside the scope. A company spokesperson said Microsoft wants staff who are committed to solving its most difficult problems. That framing presents departure as a choice for employees whose priorities no longer match the current workload.
The excluded groups are data centers, sales, cybersecurity and quantum computing, areas described as mostly important to Microsoft AI strategy. Keeping data center staff preserves capacity to build and operate the facilities behind model training and deployment. Retaining sales and cybersecurity staff protects revenue execution and the security posture around enterprise AI adoption. Quantum computing remains a longer-term research bet that management apparently does not want to interrupt.
This is the second program of this kind in 2026, following buyouts offered to 8,750 U. S. employees in May. Microsoft then cut 4,800 jobs in July, so voluntary exits are now combined with direct reductions. Total employment stands at about 223,000, roughly the same level as in 2022 despite strong revenue growth. The pattern points to reallocation rather than broad expansion, with flat headcount alongside heavier investment in systems and facilities.
What flat headcount means for AI buyers
For companies buying or building on Microsoft technology, the practical effect is continuity in AI delivery paired with turnover elsewhere. Copilot and Windows teams are inside the offer, so product support, release cadence and account contacts in those areas could change. Large enterprises with dedicated agreements will likely see less disruption than smaller customers that rely on standard support channels. Tracking response times and roadmap updates for Copilot-related products will show whether staffing changes affect delivery.
The limits of the news deserve attention because voluntary departure rates remain unknown and acceptance will vary by team and country. Inclusion in the program does not mean a role disappears, and exclusion does not guarantee faster product improvement. CFO Amy Hood says Microsoft plans to spend about 170 billion dollars on AI infrastructure in 2026, down slightly from the original 175 billion dollar plan. Buyers should therefore ask vendors which teams support their contracts, how handover is handled, and whether service levels change during the transition.
The marker to watch is the combination of final acceptance numbers and capital spending against the 170 billion dollar plan. If infrastructure outlays stay on schedule while headcount holds near 223,000, the shift toward automation-led productivity will be confirmed. A further reduction in the spending target or another workforce program would signal tighter prioritization. Either outcome will shape the pace and cost of enterprise AI offerings.
