Major US tech companies have spent a combined US$155 billion in capital expenditures so far in 2025, with much of this investment driven by the race to build AI capabilities, outpacing federal spending on education, training, employment, and social services for the year so far.

Meta, Microsoft, Amazon, and Alphabet, the parent of Google, reported tens of billions in capital expenditures in recent quarterly filings, with most of these funds directed toward building and upgrading data centers that support AI.

For the first half of the year, Meta’s capital expenditure reached US$30.7 billion, Alphabet reported nearly US$40 billion, Amazon spent US$55.7 billion, and Microsoft said it would exceed US$30 billion this quarter.

Looking ahead, Microsoft and Amazon each plan to invest about US$100 billion in capital expenditures in the next fiscal year, with AI as a major focus, while Meta expects to spend between US$66 billion and US$72 billion, and Alphabet projects US$85 billion.

Apple also increased its overall capital expenditure to US$3.46 billion last quarter, up from US$2.15 billion the previous year.

OpenAI, based in San Francisco, raised US$8.3 billion as part of a planned US$40 billion funding round.

🔗 Source: The Guardian

🧠 Food for thought

1️⃣ Massive AI investments face historically high failure rates

The $400 billion spending spree occurs against a backdrop of consistently poor AI project outcomes across the industry.

Research shows that over 80% of AI projects fail to deliver meaningful returns, with 30% of enterprise generative AI projects stalling due to poor data quality, inadequate risk controls, or unclear business value1.

This creates a disconnect between the scale of current investments and the track record of successful AI implementation.

Even with these statistics, tech companies are doubling down on infrastructure spending, suggesting they view current failures as growing pains rather than fundamental obstacles.

The pattern resembles previous technology build-outs where massive upfront investments preceded widespread adoption, though the scale of potential losses is significant in this technology cycle.

2️⃣ Infrastructure spending outpaces actual business adoption by massive margins

The tech giants are building AI infrastructure far ahead of current business demand, creating what may be one of the largest speculative technology investments in history.

Recent Microsoft developments