Google has reportedly restricted Meta's access to its powerful Gemini AI models, citing a significant shortage in available computing capacity. The move, first reported by the Financial Times, comes as Meta sought increased access to Google's AI infrastructure to power its own burgeoning AI development projects.
According to the report, Google informed Meta in March 2026 that it would be unable to provide the full computing power requested. This has led to the delay of several internal AI initiatives at Meta. The capacity crunch isn't exclusive to Meta; Google's own cloud clients have also experienced disruptions due to the high demand for AI resources. In response, Meta has advised its employees to optimize their use of AI tokens to manage the limited resources more efficiently.
Industry-Wide Compute Shortage Evident
This development underscores a broader challenge facing the technology industry: the escalating demand for AI compute capacity is rapidly outstripping supply. Despite major tech companies investing billions in AI chips, GPUs, and data centers, the infrastructure required to run advanced AI models, particularly large language models (LLMs), remains a bottleneck.
Google CEO Sundar Pichai acknowledged this constraint in a recent earnings call for Q1 2026. While Google Cloud generated $20 billion in revenue, Pichai stated that growth could have been even stronger if more computing capacity had been available.
“Obviously, we are compute-constrained in the near term,” Pichai said. “And as an example, our Cloud revenue would have been higher if we were able to meet the demand.”
The intense demand highlights the critical role of graphics processing units (GPUs) in modern AI, with companies deploying chatbots, coding assistants, and AI agents across their operations. This scarcity could open doors for other cloud providers like Amazon Web Services (AWS), Microsoft Azure, Oracle, and specialized cloud solutions to capture market share.
Beyond capacity, the rising cost of AI compute is also a growing concern. Many tech giants, including Microsoft, have reportedly begun to limit AI usage internally due to escalating operational expenses.