According to DCD, OpenAI is in talks with Amazon regarding a potential $10 billion investment agreement, as reported by The Information. The discussions, described as “very fluid,” could also see OpenAI using Amazon Web Services’ custom Trainium AI chips for its compute needs. This deal would follow a massive $38 billion multi-year compute agreement OpenAI already signed with AWS earlier this year for access to hundreds of thousands of Nvidia GB200 and GB300 chips. Should the investment proceed, it could value OpenAI at over $500 billion and potentially lead to a broader fundraising round. The news comes after HSBC analysts estimated in November 2025 that OpenAI needs around $207 billion to fund its AI data center expansion, and follows OpenAI’s confirmed move away from an exclusive cloud partnership with Microsoft in January 2025.
The Cloud Wars Just Got Hotter
Here’s the thing: this isn’t just about money. It’s about chips, power, and leverage. OpenAI signing a huge check from Amazon, while already having a $38 billion deal to use AWS’s Nvidia chips, is a massive strategic shift. Remember, Microsoft was OpenAI’s exclusive cloud partner for years. That exclusivity ended with the Stargate announcement in early 2025, and since then, OpenAI has been playing the field with Google and Oracle. But bringing Amazon into the fold as a major investor? That’s a whole new level. It turns every cloud provider into both a partner and a potential competitor. For enterprises betting on the OpenAI stack, this diversification might seem like a stability play. But it also adds a new layer of complexity. Who’s really calling the shots?
The Hardware Gambit
The real eyebrow-raiser is the potential use of AWS’s Trainium chips. OpenAI’s entire empire is built on Nvidia’s hardware. A move to also adopt Amazon’s custom silicon is a huge vote of confidence in AWS’s engineering—and a clear move to avoid being locked into a single hardware vendor. It’s the same playbook Amazon is using with Anthropic, where they’ve built “Project Rainier,” a massive cluster of hundreds of thousands of Trainium2 chips. So, Amazon would be a major investor in two of the leading AI labs. That gives them unprecedented insight and influence across the frontier AI landscape. For developers, the promise is more compute availability and potentially lower costs in the long run if competition heats up. But it also means the underlying infrastructure powering their AI models could become more fragmented.
What This Means For The AI Race
Let’s talk about that staggering number: a potential $500 billion-plus valuation. Reuters previously floated a $1 trillion IPO target. To justify those numbers, OpenAI needs to build an unthinkable amount of infrastructure—hence that $207 billion estimate from HSBC. They can’t write those checks alone. So they’re turning to the only entities with pockets deep enough: the cloud hyperscalers. This deal, if it happens, is less a traditional investment and more a strategic capital infusion to fund a specific, mind-bogglingly expensive build-out. The goal? Ensure OpenAI has the raw compute power to stay ahead of Google, Meta, and everyone else. The risk? OpenAI becomes a consortium-owned entity, pulled in different directions by its powerful benefactors at Microsoft, Amazon, and possibly others. It’s the ultimate high-stakes game. Can they take the money without losing their soul—or their strategic direction? I’m skeptical, but the financial pressure is undeniable.
