Same company. Same week. Opposite directions.
Bloomberg confirms Meta is increasing its El Paso, Texas data center investment from $1.5 billion to more than $10 billion, a more than sixfold expansion of what was already a major infrastructure commitment. CNBC reports the facility is targeting 1 gigawatt of capacity by 2028, a scale that would make it one of the largest single data center projects in North America. Meta’s 2026 capital expenditure guidance, per Yahoo Finance, sits in a range of $115 billion to $135 billion. The $135 billion figure represents the top of that guidance range, not a single confirmed number.
At the same time, NBC News reports Meta began laying off hundreds of employees across five divisions. Fox Business puts the figure at approximately 700 jobs. The Seattle Times, citing the broader context explicitly, headlined the story as cuts happening “amid record AI spending.” Total affected employees are reported as fewer than 1,000. The cuts span Reality Labs, Facebook, global operations, recruiting, and sales, a cross-functional reduction, not a single division reorganization.
The causal question is the right one to ask here. Meta hasn’t issued a public statement directly attributing the layoffs to AI investment. Multiple sources frame the events together contextually, the simultaneous timing, the scale of the infrastructure commitment, the cross-functional nature of the reductions. The cuts appear to be occurring in the context of AI infrastructure spending and restructuring. Whether the investment is directly funding the workforce reduction or the two are parallel decisions isn’t confirmed. That distinction matters for how organizations watching this interpret the signal.
For investors, the data center commitment is the primary story. A $10 billion single-facility investment, against a $115–135 billion annual capex budget, signals a company treating AI infrastructure as core operating investment rather than discretionary spending. The 1 GW capacity target by 2028 tells the technical story: Meta is building for a workload profile that doesn’t yet fully exist. It’s building ahead of demand.
For workforce planners, the layoff pattern deserves attention. Five divisions. Fewer than 1,000 total. Cross-functional. That’s a precision reduction, not a mass layoff. Whether it reflects role elimination, capability reorientation, or routine restructuring is not clear from current reporting.
What to watch: The 1 GW capacity target requires a timeline of concrete milestones – permits, construction phases, power agreements. Watch for ERCOT capacity allocations in the El Paso region, which would confirm the grid-level commitment. On the workforce side, whether Meta publishes a Worker Adjustment and Retraining Notification (WARN) filing in Texas would provide official headcount data and a clearer timeline. Follow-on quarterly earnings calls may address both the capex trajectory and any additional workforce changes.
TJS Synthesis: Meta’s week illustrates something worth tracking as a pattern across hyperscalers: the infrastructure investment cycle and the workforce reduction cycle are increasingly happening in the same reporting periods. Whether they’re causally linked or coincident depends on the company. What’s consistent is the direction, capital toward compute, fewer human roles in support functions. This brief pairs with the macro data in our AI infrastructure analysis. The pattern is larger than Meta.