Compute crunch forces Google to limit Meta's Gemini AI access after capacity request exceeds supply

Google has restricted Meta’s access to its Gemini AI models after Meta requested more computing capacity than Google could provide, according to a Financial Times report cited by multiple outlets. The shortfall, communicated around March, disrupted and delayed some of Meta’s internal AI projects, Hacker News reported. Several other Google clients have also been affected, though to a lesser extent. "Google has restricted Meta's access to its Gemini AI models due to intensifying demand for AI computing power outstripping supply, even among major tech companies." The capacity crunch highlights a broader industry challenge: AI compute shortages are now forcing even the largest technology firms to ration access to their infrastructure. Seekingalpha.com noted that Alphabet’s infrastructure bottleneck is affecting how it allocates capacity across customers. "The shortfall, communicated around March, disrupted and delayed some of Meta's internal AI projects." Both outlets pointed to the Financial Times as the original source, underscoring how quickly demand for AI computing power is outpacing available supply. While Meta’s disruption was the most significant reported, the situation illustrates that no client, however large, is immune to the constraints of Google’s data center expansion pace.

2 sources•28 JunRead the Baker's Take→