OpenAI Enters $38 Billion Agreement with Amazon

OpenAI has entered into a long-term agreement with Amazon to purchase $38 billion in AWS cloud infrastructure for the purpose of training its models and serving its customers.
This deal exemplifies the growing interconnectedness within the AI sector, positioning OpenAI at the heart of significant collaborations with major industry players such as Google, Oracle, Nvidia, and AMD.
The AWS arrangement stands out because OpenAI gained prominence partially due to its partnership with Microsoft—Amazon’s primary cloud competitor. Additionally, Amazon is a substantial supporter of one of OpenAI’s main rivals, Anthropic. Both Amazon and Microsoft are in the process of developing their own AI models to go head-to-head with startups like OpenAI.
There are rising concerns that the intense push for more infrastructure—and the atypical financial arrangements associated with these deals—may indicate the emergence of an AI bubble. Companies are anticipated to spend over $500 billion on AI infrastructure in the U.S. between 2026 and 2027, according to financial journalist Derek Thompson.
Patrick Moorhead, chief analyst at Moor Insights & Strategy, expresses that he believes leading tech firms and AI startups genuinely require additional capacity and see a feasible route to turning computing power into profits. He remarks that this new agreement illustrates that Amazon is not lagging in the AI domain after all. “Many people said they were down and out, but they just put $38 billion up on the board, right, which is pretty exceptional,” he comments.
Moorhead further notes that OpenAI’s approach is to reduce its reliance on any single cloud provider. “OpenAI is deploying with pretty much everybody at this point,” he states.
In its announcement, Amazon indicated that it is establishing customized infrastructure for OpenAI. This configuration includes two types of Nvidia chips, GB200s and GB300s, which will be utilized for both training and inference. The company also mentioned that the agreement would provide OpenAI with access to “hundreds of thousands of state-of-the-art NVIDIA GPUs, with the ability to expand to tens of millions of CPUs to rapidly scale agentic workloads.”
OpenAI and other AI entities seem to agree that agentic AI will become increasingly crucial as more users begin utilizing AI tools to navigate the online landscape.
“Scaling frontier AI requires massive, reliable compute,” stated OpenAI cofounder and CEO Sam Altman in the announcement.
Last week, OpenAI revealed plans to adopt a new for-profit structure designed to facilitate additional fundraising. While the organization remains under nonprofit control, its for-profit division has transitioned into a public-benefit corporation.
