Tech Giants Spend $700 Billion on AI Infrastructure

A group of hyperscale companies, including Alphabet Inc., Amazon, Meta Platforms, Microsoft, and Oracle, are on track to spend more than $700 billion on AI infrastructure in 2026. This massive capital expenditure has caused some investor concern regarding a potential peak in spending. However, industry analysis indicates a shift from one-time model training to recurring inference compute, which sustains demand for hardware and accelerates upgrade cycles for Nvidia products.

Investment focus is diversifying across the AI ecosystem. While Alphabet and Taiwan Semiconductor Manufacturing Company are highlighted as primary long-term holds, analysts are also identifying growth in neocloud providers. Nebius Group has secured over $20 billion in contracts from Microsoft and Meta, with a single Microsoft deal valued up to $19.4 billion running through 2031.

Further diversification is emerging in the realm of physical AI. Tesla plans to spend over $20 billion by 2026 on AI compute infrastructure and new factories to support autonomous vehicles and the Optimus humanoid robot. Despite a 10% drop in 2025 automotive revenue, the company is prioritizing long-term robotics and autonomy. Meanwhile, financial institutions like JPMorgan Chase and Goldman Sachs are directing attention toward emerging Chinese AI stocks such as MiniMax Group and Shanghai Biren Technology.