TL;DR:

• Meta has told staff it will cut roughly 8,000 jobs — 10% of its workforce — on 20 May, citing efficiency and offsetting heavy AI spending; the company will also not fill 6,000 open roles it had planned to hire into.
• Microsoft is separately offering voluntary retirement to about 7% of its US workforce, in what Bloomberg reports is the largest buyout programme in the company’s history.
• Both announcements on 23 April 2026 mark the clearest Big Tech signal yet that AI capex is being funded by headcount reallocation, not just new investment.

Meta and Microsoft have announced sweeping headcount reductions on the same day, the clearest signal yet that Big Tech firms are funding their record AI capex commitments through workforce reallocation rather than incremental investment.

The two announcements

Meta told staff via memo on Thursday that it will cut approximately 10% of its workforce — around 8,000 jobs — effective 20 May, citing efficiency and the need to offset heavy AI spending. The company also confirmed it will not hire into 6,000 open roles it had previously planned to fill, taking an additional 6,000 positions out of its future footprint. The layoffs follow meaningful headcount reductions Meta made in 2023 but are explicitly framed this time around AI funding, not efficiency alone.

Microsoft on the same day opened a voluntary retirement programme to about 7% of its US employees — the largest buyout programme in Microsoft’s history, according to a person familiar with the planning. Voluntary retirement differs from compulsory redundancy: it is employee-initiated and typically involves a lump-sum package and continued benefits. But the effect at scale is the same: reducing headcount without the morale damage of forced layoffs. Microsoft has already reduced headcount in targeted engineering groups this year, so the buyout adds a voluntary tier on top.

The pattern, read together

Against a backdrop of Intel raising AI-related revenue guidance, SAP beating cloud-growth estimates on AI-agent integration, and Computacenter flagging stronger-than-forecast data-centre demand, the Meta and Microsoft announcements draw a sharp line. The AI capex cycle is real and large, but the funding is increasingly coming from reallocating existing headcount budgets rather than net-new spend. For UK and European workforces at both companies’ regional operations, the immediate effect is limited — Meta’s cut is described as a global percentage and Microsoft’s buyout is US-only — but the message to boards and CFOs elsewhere is unambiguous.

Looking Forward

For UK enterprises modelling their own AI budgets, the Big Tech pattern matters less as a hiring signal than as a governance one: the largest AI-spending companies in the world have decided they cannot fund AI programmes purely from additional investment. That is a reality check for any UK board still treating AI deployment as incremental. Expect similar headcount-reallocation choices to surface in UK FTSE-listed firms’ next set of results, particularly in consulting, professional services and financial services where Wall Street banks have already announced AI-driven cuts.