Google restricted Meta Platforms' access to its Gemini AI models. The provider could not fulfill the full computing capacity Meta requested. Google notified Meta of the shortfall around March.
The capacity limit caused disruptions and delays to Meta's internal AI projects. This bottleneck reflects broader supply challenges affecting multiple Google Cloud clients. Meta's massive scale made the impact of these constraints particularly significant.
Technology firms are currently investing billions in data centers and specialized chips to support generative AI. These constraints at Google Cloud reveal a critical infrastructure bottleneck for the industry. The situation reinforces the strategic advantage of companies that control their own computing infrastructure.