Google caps Meta's Gemini AI usage over compute limits

Google has reportedly restricted Meta’s use of its Gemini AI models as it was demanding more AI computing power than Google could offer. According to the Financial Times report, the social media giant was reported to purchase greater access to Google's AI models to power its own AI development. However, in March, Google said that it would not be able to provide the full computing capacity.

The limited access has delayed some of Meta's internal AI projects. On the other hand, Google clients were also impacted by the capacity shortage. Now, Meta has directed its employees to efficiently use the AI tokens to limit their usage. 

Must read: Meta CPO reveals why CRED’s Kunal Shah was chosen to lead WhatsApp

While this disrupts some of Meta’s operations, it showcases the growing shortage of AI compute capacity, as demand for GPUs and large language models (LLMs) inference continues to outpace supply.

Major tech companies are spending billions of dollars on AI chips and data centres to expand their computing capacity. However, despite massive investments, they still struggle to get enough computing power to meet the demand for AI services. 

In Q1, Google Cloud generated $20 billion in revenue, showcasing strong business growth. However, the report highlighted that Google CEO Sundar Pichai said the cloud business could have grown even faster if Google had more available computing capacity. 

Must read: Why Google wants to use your old smartphone to run mini data centre

“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.” Now, this could create a way for other providers like  AWS, Azure, Oracle, and other specialised cloud providers to jump on the opportunity. 

While companies are demanding greater AI compute as they deploy chatbots, coding assistants and AI agents across their businesses. Now, they are also struggling with a greater AI bill, and many tech giants, including Microsoft, have limited AI usage amid rising AI compute bill.