Big Tech Commits $725 Billion to AI Infrastructure in 2026

Meta, Amazon, Microsoft, and Alphabet have collectively committed approximately $725 billion in capital expenditures for 2026, with the bulk of that spending directed at AI data centers, custom chips, GPUs, and AI model development. The figure represents more than a 75% increase over the prior year.

Nvidia is adding to that total with its own investments. The company is putting up to $2.1 billion into IREN for 5 gigawatts of data center capacity, underscoring the strategic importance of AI infrastructure to the chip giant.

SpaceX plans a $55 billion facility in Texas called "Terafab" for AI chip manufacturing. The facility would support Tesla's self-driving systems, humanoid robots, and AI data centers. SpaceX also secured a deal with Anthropic to provide compute capacity from its Colossus 1 data center.

OpenAI and Oracle are moving forward with a $16 billion AI infrastructure initiative in Michigan, despite local opposition. American Electric Power reported that roughly 90% of its newly contracted capacity is tied to data center customers, showing how AI demand is reshaping the power grid.

Cerebras Systems is pursuing a Nasdaq IPO to raise up to $3.5 billion, valuing the company at up to $26.6 billion. Cerebras has secured a multi-year deal with OpenAI for compute capacity.

The spending surge reflects a broader race among technology companies to secure the infrastructure needed to train and run large AI models. Venture capital is increasingly flowing toward AI infrastructure, including chips, inference engines, and data pipelines, rather than consumer-facing applications.

A report released in May 2026 warned that companies cutting staff due to AI may not achieve expected productivity gains, suggesting that AI adoption requires business redesign rather than simple cost reduction.