OpenAI has recently turned to Google for assistance with its expanding computational requirements, according to sources familiar with the matter. This decision may come as a surprise, considering the competitive relationship between the two companies in the realm of artificial intelligence. However, it underscores the significant need for computing power to facilitate the training and deployment of AI models.
The partnership between OpenAI and Google was reportedly finalized in May after extensive discussions. OpenAI has been actively seeking to reduce its reliance on Microsoft by incorporating additional cloud providers to address the growing demand for its services. This includes supporting large-scale initiatives like the Stargate project and fulfilling commitments made in recent agreements with chip manufacturers and infrastructure providers.
For Google, the collaboration with OpenAI represents an opportunity to onboard a prominent customer to its cloud platform. The company will supply additional computing resources to support OpenAI in its AI training and model operations. This development occurs amidst the ongoing competition from ChatGPT, which poses a challenge to Google’s core search business. Nevertheless, Google executives have acknowledged the potential for multiple successful entities within the AI industry.
Both OpenAI and Google, along with Microsoft, declined to comment on the specifics of the partnership. Since the launch of ChatGPT in late 2022, OpenAI has faced escalating pressure to enhance its computing capacity for training large language models and real-time inference. As of June 2025, OpenAI reported an annualized revenue of $10 billion, driven by the increased utilization of its tools.
Earlier this year, OpenAI collaborated with SoftBank and Oracle on the Stargate infrastructure endeavor, valued at $500 billion. Additionally, the company secured multi-billion-dollar agreements with CoreWeave and began developing its own chip to reduce external hardware dependencies, as reported by Reuters in February.
The agreement with Google marks a shift from OpenAI’s exclusive reliance on Microsoft’s Azure earlier this year. Although discussions between OpenAI and Google have been ongoing, existing commitments with Microsoft delayed the finalization of the deal. Microsoft and OpenAI are now revising the terms of their partnership, including Microsoft’s prospective ownership stake.
For Google, the timing of the partnership aligns with its efforts to expand the utilization of its in-house chips, known as tensor processing units (TPUs). Initially restricted to internal projects, TPUs are now accessible to external clients, including Apple and OpenAI’s competitors like Anthropic and Safe Superintelligence.
By adding OpenAI to its roster of cloud customers, Google gains another avenue for growing its cloud business, which generated $43 billion in revenue last year, constituting approximately 12% of Alphabet’s total revenue. Google has positioned itself as a neutral cloud provider for startups seeking to avoid vendor lock-in with Amazon or Microsoft, with AI companies representing a substantial segment of its clientele.
Nevertheless, Google faces its own set of challenges, including substantial investments in AI and anticipated capital spending of $75 billion this year. The company also faces regulatory scrutiny and investor pressure to deliver returns on investments.
Internally, the partnership introduces additional complexity for Google. DeepMind, Google’s AI lab, competes with OpenAI, and providing computing resources to a competitor can create tensions regarding resource allocation. CEO Sundar Pichai must balance the demands of Google’s enterprise clients with those of its AI teams.
The strain on resources has already been evident, with Google’s finance chief noting in April that the company struggled to meet cloud capacity demands in the previous quarter. Despite ChatGPT’s leadership over Gemini in user numbers, Pichai remains optimistic about OpenAI’s impact on Google’s core business.
(Photo by Jonathan Kemper)
**See also: European cloud providers play the sovereign card**
Want to learn more about cybersecurity and the cloud from industry leaders? Check out Cyber Security & Cloud Expo taking place in Amsterdam, California, and London.
Explore other upcoming enterprise technology events and webinars powered by TechForge here.
