Meta boosts annual capex sharply on superintelligence push

Meta has raised its annual capital expenditure outlook as it leans harder into what it describes as a superintelligence push. The higher spending signals that the company expects AI infrastructure, talent, and model development costs to keep climbing.

This kind of capex jump is meaningful because it turns AI ambition into balance-sheet reality. Meta is not only hiring researchers and engineers, it is spending heavily on the servers, chips, networking, and facilities needed to support a more aggressive roadmap.

The spending increase also reflects the current logic of the AI market: delaying infrastructure investments can mean falling behind. Companies that hesitate risk slower training cycles, higher serving costs, and less room to experiment with larger systems.

Meta's updated capex plan therefore reads as both offensive and defensive. It is an attempt to move faster on AI while making sure the company is not boxed out by constraints that competitors are trying just as hard to solve.