OpenAI has signed a seven-year deal worth $38 billion with Amazon’s cloud unit, Amazon Web Services (AWS). The deal will provide the AI startup with large-scale computing infrastructure, marking a huge step in the way Artificial Intelligence (AI) workloads are hosted and scaled.
The Co-founder and Chief Executive Officer (CEO) Sam Altman at OpenAI said: “Scaling frontier AI requires massive, reliable compute. Our partnership with AWS strengthens the broad compute ecosystem that will power this next era and bring advanced AI to everyone.”
The CEO of AWS, Matt Garman said: “As OpenAI continues to push the boundaries of what’s possible, AWS’s best-in-class infrastructure will serve as a backbone for their AI ambitions. The breadth and immediate availability of optimized compute demonstrates why AWS is uniquely positioned to support OpenAI’s vast AI workloads.”
Deal Overview
The strategic yearly partnership will give OpenAI access to thousands of state-of-the-art NVIDIA graphic processing units (GPUs) hosted on AWS in order to support powering ChatGPT’s real time responses and train its next-generation frontier models.
AWS has experience running large-scale AI infrastructure securely and reliably. Their leadership in cloud infrastructure combined with OpenAI’s advancements in gen AI will directly help users continue to benefit from ChatGPT.
Deployment is expected to begin immediately with full capacity by the end of 2026, and further expansion planned through 2027. The first phase of the deal involves using existing AWS data centres. The $38 billion contract is reportedly one of the largest cloud infrastructure deals in history.
The infrastructure that AWS is building clusters the NVIDIA GPUs (GB200s and GB300s) via the Amazon EC2 UltraServers on the same network which allows OpenAI to run workloads with optimal performance. The clusters have the flexibility to adapt to OpenAI’s future needs.
What’s At Stake
The scale of the deal highlights the fact that AI progress now correlates as much on computing capacity as on algorithms. Newer models like GPT5 and more will require exponentially larger resources and competition for advanced GPUs has become a defining feature of the tech landscape.
In late October, OpenAI converted its main business into a for-profit corporation, of which its longtime backer Microsoft will have a 27% stake in. The restructuring removes Microsoft’s first right to refusal to supply compute services in the new arrangement.
For OpenAI, the partnership with AWS provides a reliable, diversified supply chain for compute resources as until early this year, Microsoft was the startup’s exclusive cloud computing provider.
Strategic Impact
News of the partnership has increased Amazon shares up to 5% within hours of the announcement, which reflects investor optimism. According to Reuters, the deal could add billions in annual revenue to AWS’s books over the next decade.
OpenAI gains flexibility and leverage by aligning with AWS, as it can now expand global reach through Amazon’s data center network and secure infrastructure that is vital in maintaining its competitive edge.
AWS has emerged as the biggest winner in this deal, as it represents a vote of confidence in its infrastructure capabilities – it is currently the primary cloud provider to Anthropic rival to OpenAI. With OpenAI’s workloads running on AWS, the platform can attract other enterprise AI clients in various fields like healthcare, manufacturing or finance.
The deal could lead to competition among cloud providers like Google, Oracle, and other cloud firms which helps drive innovation and lower prices for enterprise customers. It signals a maturing phase in the market where infrastructure security becomes as strategic as model performance.
Concerns
Despite the optimism around the deal, it raises questions regarding sustainability and energy consumption as training large Ai models consumes a large amount of resources. Environmental experts have weighed in on this issue and warned that as AI workloads continue to expand, so will the carbon footprint of tech giants.
The startup is also committed to spend $1.4 trillion to develop 30 gigawatts (GW) of computing resources, that is enough to power around 25 million U.S. homes. OpenAI has a deal with Oracle to buy $300 billion in computing power for 5 years and has recently agreed to purchase $250 billion of Microsoft’s Azure cloud services as part of the restructuring deal.
The hefty commitments are being scrutinized by analysts and investors as to how the startup will fund all those deals. Though OpenAI’s revenue is expected to reach $20 billion by the end of the year, losses are also increasing as mentioned by Reuters. Since OpenAI does not make a profit, concerns over whether it can afford to pay for the infrastructure is rising.
Looking Ahead
For OpenAI, the agreement with AWS is just another step in planning to go public. By diversifying its cloud partners beyond Microsoft, OpenAI is signalling both innovation and operational resilience. Sam Altman acknowledged in a livestream recently that an IPO is the most likely path given OpenAI’s capital needs.
For Amazon, this partnership reinforces AWS’s dominance as the platform for large-scale AI workloads – positioning it to capture more enterprise demand in the years ahead. OpenAI’s foundational models are already available on Bedrock (AWS’s managed service). Companies such as Peloton and Thomson Reuters use OpenAI models for tasks like coding to agentic workflows.
Meanwhile the ripple effects are expected to reshape how cloud giants, chipmakers, and AI startups approach capacity planning and sustainability.
More to Explore
The Quest for Knowledge Continues. Fuel Your Curiosity.
