OpenAI Plans to Buy Google TPU Chips in a Deepening Challenge to Nvidia

OpenAI plans to buy Google's TPU chips as part of its massive AI infrastructure build-out, according to a report from The Information published Tuesday, a move that would deepen an unlikely alliance between two of the fiercest rivals in artificial intelligence and add fresh pressure on Nvidia's grip on the AI chip market.

The Information cited a person familiar with the matter, and Bloomberg reported the news shortly afterward. The report did not disclose financial terms or a timeline. According to Reuters, Google did not immediately respond to a request for comment outside regular business hours, while OpenAI declined to comment.

For ongoing coverage of deals like this one, AI Buzz Wire tracks every shift in the rapidly consolidating AI infrastructure market.

From Rivals to Business Partners

The reported chip purchase would mark a striking new phase in the relationship between OpenAI and Alphabet's Google. The two companies compete head-to-head in AI chatbots, frontier models, and search, but they have increasingly struck commercial deals with one another as the compute demands of frontier AI have exploded.

Just last week, OpenAI said it would make its frontier AI models available through Google Cloud, a notable concession for a company that has historically relied on Microsoft's Azure infrastructure. Reuters reported in June 2025 that OpenAI had already begun renting Google's AI chips, so Tuesday's report suggests that arrangement is graduating from an experiment into a procurement commitment.

The stakes are enormous. Amazon, Microsoft, and Google parent Alphabet plan to spend as much as $600 billion combined this year on AI infrastructure, per Reuters, as investors increasingly question when — or whether — that spending will translate into meaningful returns.

Why Google's TPUs Matter

Google's Tensor Processing Units have gained traction as one of the few credible alternatives to Nvidia's market-leading GPUs, which remain the default choice for training and serving large AI models. Google began selling TPUs directly to some customers this year, a significant shift for hardware it previously reserved almost exclusively for its own data centers.

Deals with customers including Anthropic — one of OpenAI's chief rivals — have fueled demand for Google's TPUs. Alphabet CFO Anat Ashkenazi has previously said the company expects to begin recognizing a small percentage of revenue from TPU sales agreements by the end of this year, with the vast majority expected to be recognized in 2027.

Investors will get a closer look at that momentum on Wednesday, when Alphabet reports its quarterly results, with analysts watching for signs that heavy AI investment continues to drive growth in Google Cloud.

Pressure on Nvidia

OpenAI is already among the largest buyers of Nvidia's GPUs, which it uses to train and deploy its models. A serious effort to diversify into Google silicon would nibble at the edges of a dependency that has made Nvidia one of the most valuable companies in the world.

That said, the reported plan does not signal a break with Nvidia. OpenAI has spent 2026 assembling a multi-pronged silicon strategy aimed at securing every available supply line. In June, the company unveiled its first custom AI chip designed with Broadcom, saying it planned to deploy the processor by the end of this year as part of a multi-generation chip program. OpenAI has also pursued chip manufacturing capacity through foundry partnerships as it scales, and it remains the single most important customer in Nvidia's data center business.

Buying TPUs, designing custom accelerators with Broadcom, and continuing to buy Nvidia GPUs by the fleet are not mutually exclusive strategies — they are all pieces of the same puzzle. OpenAI's challenge is that no single supplier can currently deliver the volume of compute its roadmap demands, so it is buying from everyone who can sell.

A Strange New AI Economy

What makes the report remarkable is not the procurement logic but the parties involved. Google is simultaneously OpenAI's competitor in frontier models and chatbots, its infrastructure supplier, and — through TPUs — a potential chip vendor. Similar entanglements are spreading across the industry: Anthropic buys Google's chips while competing with Google's models, and Nvidia has invested in many of its own largest customers.

The pattern reflects a hard truth of the current AI build-out. Compute is the binding constraint, and companies that need it at frontier scale cannot afford to let competitive rivalries stand in the way of securing supply.

Neither OpenAI nor Google has confirmed the reported purchase, and the terms remain unknown. But the direction of travel is clear: the boundaries between AI's competitors and its suppliers are dissolving, one multibillion-dollar deal at a time.

Stay Ahead of AI

The AI infrastructure race is reshaping the technology industry in real time. Read more AI news on AI Buzz Wire — breaking coverage of the models, chips, and deals driving the frontier.

Get the latest AI developments here.