Amazon's AI infrastructure push with Anthropic lifts outlook, but profitability debate grows

Amazon.com Inc (NASDAQ:AMZN)'s expanding artificial intelligence ambitions received a fresh vote of confidence after Anthropic agreed to spend more than $100 billion on Amazon Web Services over the next decade, prompting analysts to highlight both the scale of demand and growing scrutiny over returns.

In a note following the announcement, analysts at Jefferies said the deal, which includes plans for Anthropic to utilize up to 5 gigawatts of computing capacity, reinforces confidence in AWS's in-house chip technology, particularly its Trainium processors.

The agreement marks Amazon's second mega-scale AI infrastructure commitment in recent months, following a similar multibillion-dollar arrangement with OpenAI. Together, the deals signal what Jefferies described as "growing confidence" among leading AI developers in Amazon's custom silicon strategy.

As part of the partnership, Amazon will also invest up to $25 billion more into Anthropic, adding to its existing $8 billion stake. The cloud unit will provide access to vast computing resources, including millions of Graviton processor cores and next-generation Trainium chips.

Analysts said the scale of these commitments could support stronger long-term revenue visibility for AWS, particularly as demand for AI training and inference infrastructure accelerates.

However, the size of the deals is also expected to raise questions about capital intensity. Jefferies noted that combined commitments from Anthropic and OpenAI could account for a significant portion of the roughly 15 gigawatts of data center capacity Amazon is expected to build through 2027. That could push the company to maintain or increase its already sizable capital expenditure outlook, which had been projected at around $200 billion for 2026.

Despite concerns over rising spending, analysts pointed to potential efficiency gains from Trainium chips, which Amazon has said could lower costs significantly compared with third-party alternatives.

Jefferies said the focus now shifts to profitability, with investors likely to press management on how these large-scale AI deals translate into margins and returns. The firm described the agreements as "competitive," suggesting pricing dynamics could influence near-term profitability even as they strengthen Amazon's strategic positioning.