OpenAI is reportedly planning to purchase artificial intelligence chips from Google as technology companies expand their computing infrastructure to meet the growing demand for AI services. The development, reported by The Information and cited by Reuters on October 28, 2026, could strengthen commercial ties between the two companies, which compete in areas including AI products and search. The report did not disclose the potential deal’s financial terms, the number of chips involved or the purchase timeline. Neither company confirmed the reported plan.
The reported purchase highlights the increasing importance of specialised processors in the development and operation of artificial intelligence systems. Training large AI models and responding to user requests require substantial computing power, making access to suitable chips a key consideration for technology companies. As AI tools become more widely used for text generation, coding, research and business applications, companies are investing in the hardware and data centres needed to support these services.
Google’s Tensor Processing Units, commonly known as TPUs, are specialised chips designed to accelerate machine-learning workloads. They provide an alternative to graphics processing units, or GPUs, widely used across the AI industry. Google has developed TPUs for its own computing needs and has increasingly made the technology available to external customers. Their growing adoption reflects efforts by AI developers to expand their hardware options and potentially improve the cost and efficiency of running AI workloads.
OpenAI is already a major customer of Nvidia, whose GPUs are extensively used to train and operate AI models. Reuters previously reported that OpenAI began renting Google’s AI chips in 2025. A potential move towards purchasing TPUs could represent a further step in diversifying its computing resources, reducing reliance on a single hardware supplier and securing additional capacity for future AI services. The precise role Google’s chips would play in OpenAI’s operations has not been publicly disclosed.
The development is also significant for Google’s cloud and semiconductor businesses. Selling specialised chips and providing access to computing infrastructure can create additional commercial opportunities beyond the company’s own AI products. Demand from external customers may help Google expand its position in the AI infrastructure market, while giving businesses more options for developing and deploying machine-learning systems. Google’s ability to convert demand into sustained revenue will depend on customer adoption, available capacity and the economics of supplying the technology.
The reported plan comes amid substantial spending on AI infrastructure by major technology companies. Amazon, Microsoft and Alphabet have been preparing large investments in data centres and related infrastructure as they compete to support increasingly demanding AI workloads. This spending includes computing hardware, networking equipment, data-centre capacity and the energy infrastructure required to operate these systems. The scale of investment reflects expectations of continued demand, although technology companies also face questions about the pace at which these investments will generate returns.
Competition in AI chips is no longer limited to hardware manufacturers. Cloud providers are developing their own processors, while AI companies are exploring specialised chips tailored to their requirements. OpenAI has also developed a custom AI chip with Broadcom and has announced plans to deploy it by the end of 2026. Using a combination of external and internally developed hardware could give AI companies greater flexibility as they manage computing costs, performance and capacity requirements.
For the semiconductor industry, such developments indicate that demand for AI computing is creating opportunities across several layers of the technology supply chain. Chip designers, manufacturers, data-centre operators, cloud service providers and networking companies all play roles in delivering the infrastructure required for AI applications. As more companies seek alternatives to established GPU suppliers, competition could encourage further investment in processor design, software optimisation and specialised computing systems.
The implications extend to businesses using AI as well. Computing infrastructure influences the cost, speed and availability of AI services used for customer support, software development, analytics and automation. A wider range of hardware options could potentially improve efficiency and give service providers more flexibility in how they deliver AI products. However, the benefits for end users will depend on how successfully the hardware is integrated into software systems and whether cost savings are passed on to customers.
The reported OpenAI–Google development illustrates how commercial relationships in the technology sector can become more complex as companies compete in some markets while working together in others. Although the two companies compete in AI products, the reported chip purchase would involve Google supplying infrastructure that could support OpenAI’s operations. Such arrangements underline the importance of specialised suppliers and technology partnerships in an industry where computing capacity has become a strategic resource.
For now, the proposed purchase remains a reported plan rather than a confirmed transaction. The absence of disclosed terms means its financial value and operational impact remain unclear. If the purchase proceeds, it could further demonstrate the growing role of Google’s TPUs in the wider AI computing market and the continuing efforts of leading AI developers to secure diverse sources of processing power.
As competition in artificial intelligence intensifies, access to efficient and scalable computing infrastructure is becoming a central business priority. OpenAI’s reported interest in Google’s AI chips reflects this broader shift, in which technology companies are building more diversified hardware strategies to support their products and services. The eventual significance of the deal will depend on whether it is finalised, how the chips are deployed and what advantages they deliver in the increasingly competitive AI market.