Big Tech's Q1 Capex Tops $130 Billion, with $700 Billion AI Projection for 2026

Alphabet, Amazon, Meta, and Microsoft reported combined capital expenditures exceeding $130 billion for the first quarter of 2026, focused on AI infrastructure such as data centers and computing hardware. The companies project more than $700 billion in AI spending for the full year, compared to about $410 billion in 2025.

All four signaled sustained high levels of AI investment in their quarterly earnings released on April 29, 2026.

Meta shares fell sharply after the earnings report, dropping 7% in extended trading. The company plans to increase capital expenditure to as much as $145 billion in 2026, up from a previous maximum of $135 billion. Susan Li, Meta's chief financial officer, stated that Meta has underestimated its compute needs in past years.

He referred to Meta's Superintelligence Lab as on track to be a leading lab in the world. Meta's Hyperion data center project in northeast Louisiana costs $27 billion and is estimated to use millions of GPUs. Microsoft shares slipped after the earnings report, falling nearly 2% initially but ticking back up later.

The company's revenue increased 16% to $83 billion in the first quarter, with profits rising 23% to $38 billion. 8 billion, down almost $6 billion from a year ago. Its AI business annual run rate reached $37 billion.

Microsoft has invested more than $10 billion in its partnership with OpenAI. The company's stock is down nearly 11% so far in 2026. Alphabet shares rose after the earnings report, jumping 7% in after-hours trading due to strong cloud growth.

The company plans to spend $185 billion on AI in 2026, more than double what it spent in 2025, and signaled further increases in 2027. Alphabet's profits rose by 30% in the first quarter, and its Google Cloud business grew by 63%. 7% in extended trading due to strong cloud growth, despite an initial dip.

The company's profits increased 15% year-over-year in the first quarter, and its cloud business grew 28%. Amazon's annual run rate for its own AI chips is $20 billion.

Amazon has laid off more than 30,000 workers in recent months. A single Nvidia GPU can cost up to $40,000, an eight-GPU server can cost hundreds of thousands of dollars, and clusters for hyperscale AI data centers, made up of thousands or hundreds of thousands of GPUs, can cost billions of dollars.

7 trillion worldwide by 2030. , targeting a valuation of up to $100 billion, according to the Financial Times. The IPO could occur as early as the second half of 2026.

SoftBank's Roze venture focuses on physical buildout of AI infrastructure using robotics to construct data centers. SoftBank has invested tens of billions in OpenAI and other large-scale infrastructure projects. 1.

OpenAI acknowledged the goblin-referencing behavior in a blog post on April 29, 2026. The behavior persisted in later models like Codex after the Nerdy personality was removed. Brian X. Chen tested Google's Gemini AI as a trip planner in a New York Times column.

Google's Gemini AI integrates with Google services including Flights, Hotels, Gmail, and Maps. 75% of tech leaders agree that their operating models and processes need to change in the next 12 to 18 months to drive greater value from AI, according to Deloitte's 2026 Global Tech Leadership Study.

80% of tech leaders are confident in their organization's ability to deploy and govern AI capabilities at scale, according to the same study.

Alphabet, Amazon, Meta, and Microsoft all beat first-quarter earnings expectations on April 29, 2026.