Amazon AWS AI Revenue Tops $15B as Jassy Defends Spending

Amazon Breaks Out AI Revenue for the First Time

Amazon made its most significant financial disclosure about artificial intelligence to date on April 9, 2026, revealing that AWS's AI revenue run rate exceeded $15 billion in the first quarter. The figure, shared in CEO Andy Jassy's annual shareholder letter, marks the first time Amazon has isolated AI-specific revenue from its broader cloud business and provides one of the clearest signals yet that hyperscaler AI spending is translating into real top-line growth.

The $15 billion AI run rate represents roughly 10% of AWS's total $142 billion revenue run rate, and Jassy emphasized that the trajectory is accelerating. He described customer demand as outstripping available capacity, suggesting AWS would be growing even faster without infrastructure constraints.

"These numbers are ascending rapidly. If anything, we're capacity-constrained — demand for AI services continues to outpace what we can provision." — Andy Jassy, Amazon CEO

Custom Chips Business Doubles to $20 Billion

In the same letter, Jassy disclosed that Amazon's custom silicon business — which produces Graviton general-purpose processors and Trainium AI training chips — now generates an annual revenue run rate above $20 billion. That figure has doubled from the $10 billion milestone the company cited earlier in 2026, underscoring Amazon's aggressive push to reduce dependency on Nvidia GPUs.

Amazon's Trainium2 chips, which began shipping to large customers in late 2025, are being positioned as a cost-effective alternative to Nvidia's H100 and B200 accelerators. Anthropic, one of AWS's highest-profile AI customers, is among the companies training models on Trainium infrastructure as part of a multi-billion-dollar partnership.

Jassy Defends $200 Billion Capex Plan

The revenue disclosures arrive as Amazon faces investor scrutiny over its massive capital expenditure plans. In February, the company committed to spending approximately $200 billion in capex during 2026, primarily on AI infrastructure including data centers, networking equipment, and custom chip fabrication. Jassy's letter was clearly crafted to demonstrate that the spending is producing returns.

The letter drew parallels to AWS's early years, when heavy infrastructure investment preceded years of compounding growth. Jassy argued that AI represents a similar inflection point, with enterprise adoption still in the early stages despite the revenue acceleration.

What This Means for Engineers and Job Seekers

For cloud engineers and AI infrastructure specialists, Amazon's disclosures signal continued aggressive hiring across its AI divisions. Roles involving Trainium optimization, custom chip design, and large-scale distributed training are likely to expand as Amazon races to meet demand. For job seekers, AWS's AI growth translates to opportunities not just at Amazon but across the thousands of companies building on AWS AI services — from startups using Bedrock to enterprises deploying SageMaker at scale.