For the first three years of the artificial intelligence boom, Amazon.com Inc. watched as cloud customers seeking the latest and greatest from OpenAI took their business to Microsoft Corp.

But the day after Amazon’s biggest cloud rival agreed to drop its exclusive rights to resell OpenAI products, Amazon says it will finally be able to make the ChatGPT maker’s AI models available to its own customers. 

“It’s something that our customers have asked for, for a really long time,” Matt Garman, chief executive officer of the Amazon Web Services cloud unit, said in an interview with Bloomberg Television. 

Some of OpenAI’s latest models will be available to preview on AWS starting Tuesday, with the most powerful GPT models arriving “in the next couple of weeks,” Garman said. 

Microsoft got a jump on the large language model era thanks to a massive early investment in OpenAI. That gave Microsoft’s Azure cloud unit exclusive rights to sell OpenAI’s most powerful models.

After the late 2022 launch of ChatGPT, Amazon scrambled to assemble a portfolio of the best of the rest on its Bedrock AI model marketplace, including products from OpenAI archrival Anthropic PBC and Meta Platforms Inc. Still, some longtime AWS customers, including Superhuman, formerly Grammarly, turned to Microsoft for AI services. 

Amazon invested $50 billion in OpenAI earlier this year, by far its biggest investment in any business, and OpenAI said it would spend an additional $100 billion on AWS computing power and chips. 

“We absolutely are expecting a lot of growth there,” Garman said. 

The two companies have jointly developed a product based on OpenAI’s models to help autonomous AI agents understand context and remember prior interactions with users. That tool, called Amazon Bedrock Managed Agents, was introduced on Tuesday at an event in San Francisco where AWS also announced a bigger push into business applications.

Amazon’s tighter relationship with OpenAI coincides with growing worries that the richly valued startup won’t expand its sales fast enough to meet commitments it’s making to Amazon and other infrastructure providers, including Microsoft and Oracle Corp. 

On Monday, the Wall Street Journal reported that OpenAI had fallen short of several internal targets. OpenAI on Tuesday pushed back, saying it was “firing on all cylinders.” 

AWS’s Garman said there was still more demand for AI services than supply of computing power to meet it. 

“The OpenAI team, I think, would gladly take more capacity from us this year and next year and the year after that as we add it,” he said. “And so we’re still working really hard to continue to add more capacity, whether it’s power, whether it’s chips, whether it’s memory, all around the world.”

More stories like this are available on bloomberg.com

©2026 Bloomberg L.P.

Published on April 29, 2026