OpenAI Signs $38 Billion, Seven-Year Cloud Deal with Amazon to Power Next Phase of AI Expansion

OpenAI has signed a landmark seven-year, $38 billion cloud-computing agreement with Amazon.com Inc., marking its first major strategic partnership since a recent corporate restructuring that gave the ChatGPT maker greater autonomy over its operations and finances. The deal represents one of the largest cloud infrastructure commitments in history and signals OpenAI’s next phase of expansion as it races to build increasingly powerful artificial intelligence models.

Under the agreement, OpenAI will use Amazon Web Services (AWS) to train and deploy its next generation of large language models, tapping into hundreds of thousands of Nvidia graphics processors housed across Amazon’s global data centers. The companies said the new compute capacity would come online gradually through 2026, with room for additional expansion in the years that follow.

The deal highlights the AI industry’s voracious appetite for compute power as companies escalate their pursuit of artificial general intelligence (AGI). OpenAI Chief Executive Sam Altman said that scaling up compute is fundamental to advancing AI capabilities. “Scaling frontier AI requires massive, reliable compute,” Altman said in a statement. “Our partnership with AWS strengthens the broad compute ecosystem that will power this next era and bring advanced AI to everyone.”

The partnership also represents a crucial win for Amazon Web Services, which has been under pressure to demonstrate its competitiveness against Microsoft Azure and Google Cloud in the rapidly growing AI market. AWS shares surged to an all-time high following the announcement, lifting Amazon’s market capitalization by nearly $140 billion. Analysts said the agreement reaffirmed AWS’s position as a global leader in hyperscale cloud infrastructure capable of supporting frontier AI workloads.

“This is a hugely significant deal and a clear endorsement of AWS’s compute capabilities to deliver the scale needed to support OpenAI,” said Paolo Pescatore, analyst at PP Foresight. “It underscores that even the most advanced AI companies continue to rely on hyperscalers for the infrastructure backbone of their innovation.”

Amazon plans to deploy Nvidia’s latest GB200 and GB300 AI accelerators within massive data clusters that will power ChatGPT’s responses and model training operations. The new systems will serve as part of a broader push by OpenAI to create an estimated 30 gigawatts of computing capacity—an investment Altman has previously valued at $1.4 trillion, roughly equivalent to the energy demand of 25 million U.S. homes.

The partnership comes just weeks after OpenAI finalized a restructuring that loosened Microsoft’s preferential compute-supply rights, giving the San Francisco-based firm more flexibility to work with other cloud providers. The company is also reportedly preparing for an initial public offering that could value it at up to $1 trillion, according to Reuters.

While the agreement deepens OpenAI’s cloud footprint, it also amplifies broader market concerns over the soaring costs of AI infrastructure and the risk of overheating investment in the sector. With more than $1 trillion already committed by OpenAI and its backers to compute expansion, analysts warn that the AI industry’s capital intensity could mirror the speculative dynamics of previous tech bubbles.

Nevertheless, for both Amazon and OpenAI, the deal represents a powerful alignment of interests—one seeking to cement cloud leadership, the other to fuel the next great leap in artificial intelligence.