Amazon to raise at least $25 billion in bonds to fund AI infrastructure

On July 7, 2026, Amazon announced plans to sell at least $25 billion in dollar-denominated bonds on the US market, one of the largest corporate debt offerings of 2026. The proceeds will fund large-scale investment in AI infrastructure.

Why does Amazon need $25 billion?

The company stated the purpose directly: financing AI infrastructure spending. That covers the entire technology stack that supports AI workloads: building and expanding data centers, buying specialized AI accelerators, and developing software platforms for training large models and running inference at the scale of millions of users.

A dollar bond issue is a standard tool for corporations with high credit ratings. The company borrows on the open market at a fixed rate without diluting shareholders, and channels the funds into capital-intensive projects with long payback horizons. Modern AI data centers fit that description exactly, with construction costs running to billions of dollars per site.

Key terms of the deal:

• minimum size: $25 billion
• format: dollar bonds on the US market
• purpose: financing AI infrastructure spending
• Bloomberg calls it the "latest stage" of Amazon's growing AI investments

How Amazon is building its AI stack

Amazon Web Services, the world's largest cloud provider by revenue, is locked in fierce competition with Microsoft Azure and Google Cloud for enterprise customers' AI workloads. Demand has surged: businesses are moving model training, generative AI services and complex analytics to the cloud, choosing providers on price, latency and availability of specialized compute.

Amazon is building its AI infrastructure along several lines:

• In-house AI chips: the Trainium (training) and Inferentia (inference) series reduce reliance on NVIDIA GPUs and lower compute costs for AWS customers.
• Amazon Bedrock: a managed platform through which thousands of companies run third-party AI models, including Claude from Anthropic.
• A global data center network: Amazon is expanding facilities worldwide to keep AI service latency low.
• Investment in Anthropic: multibillion-dollar stakes in the developer of one of the leading AI model families.

All of these lines require steadily rising capital expenditure, which directly explains the size of the current offering.

Why does the race demand ever more money?

Demand for AI compute is growing faster than the supply of infrastructure. New generations of language and multimodal models need far more resources for training and inference than their predecessors did a few years ago. Enterprise demand is rising in parallel, as companies build AI into products and internal processes.

In these conditions, tech giants are placing bets on infrastructure years before demand fully shows up in revenue. The debt market lets them fund those bets while preserving operational flexibility.

What this means

An offering of at least $25 billion signals to the market that Amazon does not intend to slow its AI investment and is ready to raise large amounts of capital to hold its lead in cloud and AI. For the industry, it sets a benchmark for scale: the AI infrastructure race in 2026 is measured in tens of billions of dollars, and each new offering raises the bar for competitors.