OpenAI has announced it will use Google’s cloud infrastructure to power ChatGPT and its programming interface across several countries, marking a significant expansion beyond its initial reliance on Microsoft’s cloud services.
The artificial intelligence company said ChatGPT will run on Google Cloud Platform alongside existing suppliers Microsoft, CoreWeave, and Oracle. The Google infrastructure will operate in the United States, Japan, the Netherlands, Norway, and the United Kingdom.
The move represents OpenAI’s continued effort to secure additional computing power to meet heavy demand for its AI services. The company initially relied exclusively on Microsoft for cloud capacity but has since diversified its infrastructure partnerships.
OpenAI CEO Sam Altman highlighted the company’s capacity challenges in April, writing on social media: “if anyone has GPU capacity in 100k chunks we can get asap please call!” The comment underscored the startup’s need for more graphics processing units to power its large language models.
The partnership marks a notable win for Google’s cloud division, which remains smaller and younger than competitors Amazon Web Services and Microsoft Azure. Google Cloud has also secured business with Anthropic, an AI company founded by former OpenAI executives.
Relations between OpenAI and Microsoft have evolved since their initial exclusive partnership. Microsoft named OpenAI as a competitor last year, though the companies continue to collaborate. Both firms now sell competing AI tools for developers and offer business subscriptions.
In January, Microsoft agreed to provide OpenAI with a “right of first refusal” for additional computing resources rather than serving as its exclusive vendor. However, Microsoft retains exclusive rights to OpenAI’s programming interfaces.
OpenAI has significantly expanded its cloud partnerships over the past year. In March, the company signed a five-year deal with CoreWeave worth nearly $12 billion. Last year, Oracle announced it was partnering with Microsoft and OpenAI to extend Microsoft’s AI platform to Oracle’s cloud infrastructure.
The Google partnership was first reported by Reuters in June, with Wednesday’s announcement providing official confirmation and additional details about the geographic scope.
The expansion reflects the enormous computing demands of modern AI systems, which require vast amounts of processing power to train and operate large language models like those powering ChatGPT.
