Meta faces limits in using Google’s Gemini AI due to insufficient computing power, reveals Financial Times

Technology giant Google has established restrictions on the use of its Gemini artificial intelligence models by Meta, according to a report in the Financial Times. The measure was taken due to Google’s inability to offer all of the computing capacity requested by the social media company.
This episode highlights the growing difficulties facing artificial intelligence infrastructure. Alphabet, Google’s parent company, applied these limitations to several customers, with Meta feeling a significant impact, the FT detailed on Sunday. This decision resulted in a ripple effect on Meta’s internal plans, leading the company to instruct its employees to optimize the use of AI tokens, according to anonymous sources consulted by the newspaper.
When contacted by the Financial Times for comments on the matter, both Google and Meta chose not to comment.
Initially, Meta used Gemini AI, recognized for its superior performance to the Llama open source models developed by the company itself. It employed technology to automate security tasks such as removing harmful content and preventing fraud. However, the company has intensified the use of its internal model, Muse Spark, aiming to reduce dependence on third-party artificial intelligence solutions.
The rapid growth of artificial intelligence has put pressure on computing capacity and energy demand. This situation directly challenges the expansion of the countless data centers that are under construction around the world.
In a strategic move at the beginning of June, Google signed an agreement with Elon Musk’s SpaceX, committing to pay US$920 million per month for processing capacity. This contract, valued at 30 billion dollars in cloud computing services, extends until mid-2029 and demonstrates Google’s urgency in meeting the immense demand for AI capacity.
Earlier this year, Meta communicated to its employees its intention to reduce its workforce by 10%, which is equivalent to approximately 8,000 positions. This action aimed to compensate for high investments in artificial intelligence. At the same time, the company reassigned 7,000 employees to new AI-focused roles as part of a comprehensive corporate restructuring.






