OpenAI, the firm behind ChatGPT, has struck a massive seven-year, $38 billion agreement with Amazon Web Services (AWS) to secure the cloud compute capacity it needs to fuel its relentless AI ambitions. This landmark deal is the first major move by OpenAI since its corporate restructuring last week, which gave the AI giant crucial operational and financial independence.
The timing of the deal announcement, made on Monday, is critical. OpenAI recently reorganized its core business, a move that included removing Microsoft’s first right of refusal to be its sole compute provider. This strategic realignment paved the way for the Amazon partnership, marking OpenAI’s transition to a multi-cloud infrastructure strategy.
This diversification guarantees access to the immense, reliable compute power necessary for exponential growth, mitigating reliance on any single partner.
A vote of confidence for AWS
For AWS, the contract serves as a major vote of confidence, silencing investors who feared the e-commerce giant’s cloud unit had fallen behind rivals like Microsoft and Google in the AI race.
The core of the $38 billion deal is dedicated to securing the muscle for next-generation AI development.
Amazon plans to roll out hundreds of thousands of chips, including Nvidia’s GB200 and GB300 AI accelerators, in data clusters built to power ChatGPT’s responses and train OpenAI’s next wave of models, the companies said.
“This is a hugely significant deal (and is) clearly a strong endorsement of AWS compute capabilities to deliver the scale needed to support OpenAI,” PP Foresight analyst Paolo Pescatore told Reuters.
Amazon already offers OpenAI’s open-weight models on Amazon Bedrock, which provides multiple AI models for businesses using AWS.
This hardware is essential for two key functions:
- Training OpenAI’s next wave of frontier models.
- Powering the real-time responses for consumer-facing applications like ChatGPT.
OpenAI is set to begin using AWS immediately, with the full planned capacity coming online by the end of 2026. The terms also include flexibility for further expansion in 2027 and beyond.
Sam Altman’s trillion-dollar vision
The astronomical size of the AWS commitment is merely a stepping stone toward the vision of CEO Sam Altman. He has publicly stated that the startup is determined to spend an estimated $1.4 trillion to build 30 gigawatts of computing resources—an energy commitment roughly equivalent to powering 25 million US homes.
Altman’s long-term goal is to eventually add 1 gigawatt of compute every week, a figure he admits comes with an eye-watering capital cost of over $40 billion per gigawatt.
“Scaling frontier AI requires massive, reliable compute,” said Altman. “Our partnership with AWS strengthens the broad compute ecosystem that will power this next era and bring advanced AI to everyone.”
How the market reacted
The news sent a seismic wave through the stock market. Amazon shares jumped 5% on Monday, hitting an all-time high and adding nearly $140 billion to its market value. Conversely, Microsoft shares briefly dipped on the news.
This deal firmly cements AWS as a foundational backbone for the world’s leading AI companies.
