Google caps Meta’s Gemini access – compute scarcity is now a real constraint for AI-dependent enterprises

Google told Meta around March that it could not fulfil the full Gemini compute
capacity Meta had sought to purchase. The shortfall has disrupted and delayed
some of Meta’s internal AI projects, with staff now being asked to use AI tokens more
efficiently. Other Google clients were affected too, but Meta bore the biggest impact
due to exceptionally high demand. The report, first published by the Financial Times,
was not confirmed by either company.

If Google is rationing capacity to Meta, smaller enterprises are at greater risk of
supply constraints. Enterprises running AI-critical workloads on a single cloud
provider should assess multi-cloud or open-source fallback strategies now, before
demand-driven rationing becomes a recurring reality.