The news: OpenAI struck a landmark $300 billion deal with Oracle to build AI data centers across the US, cementing Oracle as a critical partner in the race to scale artificial intelligence. The agreement, part of Project Stargate, covers more than half of the computing infrastructure OpenAI says it will need over the next five years, per The New York Times.

News of the deal sent Oracle shares up 40% in a day, boosting co-founder Larry Ellison’s fortune by $100 billion and briefly making him the world’s richest person, ahead of Elon Musk.

Why it’s worth watching: Oracle is quickly becoming a central force in the AI cloud era. Fueled by mega deals with OpenAI, Meta, and Nvidia—and its role as a key partner in Project Stargate—Oracle’s AI infrastructure pivot is creating new opportunities for growth.

Oracle has excelled at inking partnerships with the biggest players in generative AI (genAI). 

• Oracle and Nvidia AI Enterprise are jointly negotiating sovereign AI clouds for various markets. 
• The company also committed $8 billion to Japan’s cloud computing and AI infrastructure over the next 10 years.

Its large install base of 430,000 customers in 175 countries demonstrates an ability to scale while locking down security and compliance risks for enterprises.

Oracle’s second act? Oracle’s pivot shows legacy tech brands can reinvent themselves by leveraging trust, infrastructure, and enterprise credibility rather than competing directly with hyperscalers on consumer AI. Continued success puts Oracle among Big Tech’s most influential companies. 

Its resurgence is beating expectations, considering only 13% of US business leaders in 2024 expected it to be a genAI leader in three to five years, per Wharton School of Business.

Yes, but: Oracle’s backlog, which is the pile of future orders it has signed but not yet delivered or collected payment for, tells a different story. It jumped 359% YoY, from $138 billion last quarter to $455 billion. 

That looks like massive demand, but backlog isn’t cash in the bank—many of these contracts may take years to turn into revenues, and some may never fully materialize.

• A swollen backlog can mask execution risk: Delays in data center builds or AI adoption slowdowns could push revenues out further.
• Customers may renegotiate or cancel portions if market conditions shift, leaving Oracle exposed.

Our take: AI’s future rests on who can actually deliver compute at scale. Marketers should diversify cloud and AI partners, experiment early, and prepare to shift strategies quickly as winners and losers emerge in this infrastructure race.

Oracle’s opportunity is enormous—but so is the execution risk.