Meta Moves to Share AI Costs, Sells $2 Billion in Data Center Assets

Meta Moves to Share AI Costs, Sells $2 Billion in Data Center Assets

Written by Mark Williams, In News, Published On
August 2, 2025
, 15 Views

Meta Platforms Inc. (META.O) has said it will sell data centre assets worth $2 billion as it wants to work with outside partners to pay for its massive AI infrastructure. IT companies used to pay for their growth. Still, now they are looking for new ways to deal with the high costs of making creative AI technologies.

This is a significant change.

Meta’s most recent quarterly report said that the company wants to change the status of $2.04 billion worth of land and buildings to “held-for-sale.” You can now give money to third-party partners to help build data centres with you next year, thanks to this change. Meta’s Chief Financial Officer, Susan Li, said that the company will continue to pay for most of its capital investments with money that it already has on hand. But some projects might be able to get money from outside sources. This would give the company more choices as the needs for AI systems change.

CEO of Facebook, Mark Zuckerberg, has been very open about his big plans to build “superclusters” for superintelligence, which are artificial intelligence data centres. He may have to spend hundreds of billions of dollars to make these come true.  These superclusters are important to Meta’s long-term AI goal; one of them is said to cover a lot of Manhattan. The company is still spending more on equipment, but its efforts to find financial partners should help bring these costs down.

According to Meta, “capital spending will reach $66 billion to $72 billion this year, which is $2 billion more than it was before”.

Meta also quoted that the ad sales were better than expected, and all thanks to AI-driven improvements in tracking and delivering content. Even though the company had to spend a lot of money, this helped make up for rising building costs. The way Meta does things is part of a bigger trend in the tech industry to work together and make things together, even if AI costs a lot.

Related articles
Join the discussion!