Google is restricting Meta Platforms’ access to Google’s Gemini artificial intelligence models because it cannot supply the full amount of computing capacity Meta requested, the Financial Times reports. The report says Google told Meta around March that it could not meet the complete Gemini capacity Meta sought to purchase, leading to limits on how much Gemini compute and related access Meta can use. Several other Google customers are also affected, though Meta appears to experience the biggest impact because it has unusually high demand for the models, according to the FT.
The capacity constraints disrupt and delay some of Meta’s internal AI projects, the report adds. It also says Google has encouraged Meta staff to make more efficient use of “AI tokens,” which measure AI usage, after the restrictions were put in place.
More broadly, the articles note that even as companies invest heavily in chips and data centres, they continue to face shortages in computing power to meet growing demand for AI services. The Hindu also links the issue to Google Cloud’s performance and remarks by Google CEO Sundar Pichai about computing power constraints contributing to higher backlog and limiting growth.
