Google caps Meta’s use of its Gemini AI models: Report

Google has placed limits on Meta’s use of its Gemini artificial intelligence models after the Facebook parent sought more computing capacity than the company could provide, according to a report by the Financial Times, highlighting the growing strain on AI infrastructure despite record investments by major technology firms.

Google informed Meta around March that it could not provide the full Gemini computing capacity the social media company had sought to purchase, the Financial Times reported, citing people familiar with the matter. The shortfall disrupted and delayed some of Meta’s internal AI projects, according to the report.

As the restrictions remain in place, Meta has also asked employees to use AI tokens (the units that measure AI usage) more efficiently as part of a broader effort to optimise computing resources and AI spending.

The news report said Meta has been affected more than other Google Cloud customers because of its exceptionally high demand for Gemini models. Other clients have also experienced capacity constraints, although to a lesser extent.

AI boom continues to strain computing capacity

Alphabet has previously acknowledged these constraints. During its first-quarter earnings, Chief Executive Sundar Pichai said demand for AI services exceeded available computing capacity, limiting Google Cloud’s growth despite strong customer demand. Google Cloud generated $20 billion in revenue during the quarter, while capacity shortages contributed to a sharp increase in its order backlog.

According to the Financial Times, Meta has increasingly relied on Google’s Gemini models for workloads including software development, safety automation and customer service, even as it invests heavily in its own AI capabilities. The company is also working to reduce its dependence on external providers by expanding the use of its in-house AI models, the report said.

As per a McKinsey report, their research shows that by 2030, data centres are projected to require $6.7 trillion worldwide to keep pace with the demand for compute power. Data centres equipped to handle AI processing loads are projected to require $5.2 trillion in capital expenditures, while those powering traditional IT applications are projected to require $1.5 trillion in capital expenditures.