Meta's $2 Billion Asset Sale Signals Shift in AI Infrastructure Funding Strategy

Meta to share AI infrastructure costs via $2 billion asset sale

Aug 1 (Reuters) - Meta Platforms (META.O), opens new tab is pressing ahead with efforts to bring in outside partners to help fund the massive infrastructure needed to power artificial intelligence, disclosing plans in a filing on Thursday to offload $2 billion in data center assets as part of that strategy. The strategy reflects a broader shift among tech giants -- long known for self-funding growth -- as they grapple with the soaring cost of building and powering data centers to support generative AI. The social media giant said earlier this week that it was exploring ways to work with financial partners to co-develop data centers to help finance its massive capital outlay for next year. "We're exploring ways to work with financial partners to co-develop data centers," Meta Chief Finance Officer Susan Li said on a post-earnings conference call on Wednesday. While the company still expects to fund much of its capital spending internally, some projects could attract "significant external financing" and offer more flexibility if infrastructure needs shift over time, Li said. The company did not have any finalized transactions to announce, she said. The disclosure in Meta's quarterly filing, however, signals that plans are firming up. In its quarterly filing on Thursday, Meta said it had approved a plan in June to dispose of certain data center assets and reclassified $2.04 billion worth of land and construction-in-progress as "held-for-sale". These assets were expected to be contributed to a third party within the next twelve months for co-developing data centers. Meta did not record a loss on the reclassification, which values the assets at the lower of their carrying amounts or fair value less costs to sell. As of June 30, total held-for-sale assets stood at $3.26 billion, according to the filing. Meta declined to comment for this story. CEO Mark Zuckerberg has laid out plans to invest hundreds of billions of dollars into constructing AI data center "superclusters" for superintelligence. "Just one of these covers a significant part of the footprint of Manhattan," he said. The Instagram and WhatsApp owner on Wednesday raised the bottom end of its annual capital expenditures forecast by $2 billion, to $66 billion to $72 billion. It reported stronger-than-expected ad sales, boosted by AI-driven improvements to targeting and content delivery. Executives said those gains were helping offset rising infrastructure costs tied to its long-term AI push. Reporting by Echo Wang in New York; Editing by Sayantani Ghosh and Marguerita Choy Our Standards: The Thomson Reuters Trust Principles., opens new tab * Suggested Topics: * Artificial Intelligence * Grid & Infrastructure Echo Wang Thomson Reuters Echo Wang is a correspondent at Reuters covering U.S. equity capital markets, and the intersection of Chinese business in the U.S, breaking news from U.S. crackdown on TikTok and Grindr, to restrictions Chinese companies face in listing in New York. She was the Reuters' Reporter of the Year in 2020.

Meta pledges to spend billions on AI infrastructure - but will this help it catch up in the AI race

As governments and private companies around the world announce huge AI infrastructure plans, Meta has become the latest to announce a major spending spree. In its most recent quarterly financial results, the company announced an expansion of spending to around $66-$72 billion - more than doubling its spending on data centres and servers. "We expect that developing leading AI infrastructure will be a core advantage in developing the best AI models and product experiences, so we expect to ramp our investments significantly in 2026 to support that work," Susan Li, Meta CFO, said during the company's earnings call. The move comes shortly after Meta's shares jumped more than 10%, thanks largely to advertising business success, where AI-driven ad creation tools allow users to generate video ads from images, helping prompt better conversion rates. Meta says it expects a similar jump in spending onwards in 2026, as the firm looks to "aggressively [pursue] opportunities to bring additional capacity online to meet the needs of [its] artificial intelligence efforts and business operations." Meta isn't alone, as Microsoft has also announced plans to spend billions on AI, announcing over $30 billion capital investment as it expands AI capacity in a bid to keep up with the likes of Amazon - which itself is set to spend up to $111 billion in 2025, with the majority going towards tech and infrastructure. This infrastructure, primarily referring to data centres and large servers, is costly in more ways than one. It would be disingenuous to talk about data centre spending without mentioning the enormous environmental costs associated with the infrastructure. Data centres consume massive amounts of energy and water - depleting local water sources and putting huge strain on already labored energy grids. Local communities are being hit hard by nearby data centre constructions - with data centres in Texas using 463 million gallons of water, as residents are instructed to take shorter showers to offset the usage. In Georgia, residents living in proximity to Meta's own data centre, can no longer drink their water, with taps running dry thanks to added sediment in local wells. The cost of municipal water has skyrocketed, and the county water commission may face a shortage.

Meta to spend up to $72B on AI infrastructure in 2025

Meta Platforms is significantly increasing its investment in artificial intelligence infrastructure, encompassing data centers and servers, as revealed in its second-quarter earnings report. The company projects capital expenditures for 2025 to reach between $66 billion and $72 billion. This projection represents an increase of approximately $30 billion year-over-year at the midpoint. Meta anticipates a similarly substantial surge in spending on AI infrastructure in 2026. This continuous investment aims to bolster the company's capacity to meet the evolving demands of its artificial intelligence initiatives and broader business operations. Susan Li, Meta CFO, emphasized the strategic importance of AI infrastructure during the company's earnings call. She stated, "We expect that developing leading AI infrastructure will be a core advantage in developing the best AI models and product experiences, so we expect to ramp our investments significantly in 2026 to support that work." While Meta intends to primarily fund its AI endeavors independently, the company is exploring collaborative opportunities with financial partners for the co-development of data centers. Li mentioned, "We don't have any finalized transactions to announce, but we generally believe that there will be models here that will attract significant external financing to support large-scale data center projects that are developed using our ability to build world-class infrastructure, while providing us with flexibility should our infrastructure requirements change over time." Meta has unveiled plans for two major AI "titan clusters." Prometheus, located in Ohio, is projected to be among the first AI superclusters to achieve 1 gigawatt of compute power upon its anticipated launch in 2026. Additionally, Hyperion, situated in Louisiana, is envisioned by Meta CEO Mark Zuckerberg to potentially scale up to 5 gigawatts over the coming years, encompassing a footprint comparable in size to Manhattan. Furthermore, Meta is actively developing several other titan-scale clusters, though specific details remain undisclosed. These extensive data center projects are expected to consume substantial amounts of energy, potentially impacting nearby communities. One Meta project in Newton County, Georgia, has reportedly led to water shortages for some residents. Meta's earnings report also highlighted employee compensation as the second-largest driver of growth. The company is investing heavily to attract skilled AI engineers and researchers to its newly established Superintelligence Labs business unit. Prior to the earnings announcement, Zuckerberg outlined his vision for "personal superintelligence," emphasizing AI's potential to enhance individual lives, particularly through Meta's smart glasses and virtual reality headsets. Following the release of the earnings report, Meta's stock experienced a 10% surge in after-hours trading, reflecting investor confidence in the company's overall performance and optimistic outlook for the third quarter. Meta reported revenue of $47.5 billion for the second quarter, and anticipates revenue between $47.5 billion and $50.5 billion in the third quarter. Advertising revenue, boosted by AI-powered tools, played a significant role in Meta's financial gains. These tools facilitate AI-powered translations and video generation, enabling advertisers to create more effective and targeted campaigns.

Meta to share AI infrastructure costs via $2 billion asset sale - The Economic Times

In its quarterly filing on Thursday, Meta said it had approved a plan in June to dispose of certain data center assets and reclassified $2.04 billion worth of land and construction-in-progress as "held-for-sale". These assets were expected to be contributed to a third party within the next twelve months for co-developing data centers.Meta Platforms is pressing ahead with efforts to bring in outside partners to help fund the massive infrastructure needed to power artificial intelligence, disclosing plans in a filing on Thursday to offload $2 billion in data center assets as part of that strategy. The strategy reflects a broader shift among tech giants - long known for self-funding growth - as they grapple with the soaring cost of building and powering data centers to support generative AI. The social media giant said earlier this week that it was exploring ways to work with financial partners to co-develop data centers to help finance its massive capital outlay for next year. "We're exploring ways to work with financial partners to co-develop data centers," Meta Chief Finance Officer Susan Li said on a post-earnings conference call on Wednesday. While the company still expects to fund much of its capital spending internally, some projects could attract "significant external financing" and offer more flexibility if infrastructure needs shift over time, Li said. The company did not have any finalised transactions to announce, she said. The disclosure in Meta's quarterly filing, however, signals that plans are firming up. In its quarterly filing on Thursday, Meta said it had approved a plan in June to dispose of certain data center assets and reclassified $2.04 billion worth of land and construction-in-progress as "held-for-sale". These assets were expected to be contributed to a third party within the next twelve months for co-developing data centers. Meta did not record a loss on the reclassification, which values the assets at the lower of their carrying amounts or fair value less costs to sell. As of June 30, total held-for-sale assets stood at $3.26 billion, according to the filing. Meta declined to comment for this story. CEO Mark Zuckerberg has laid out plans to invest hundreds of billions of dollars into constructing AI data center "superclusters" for superintelligence. "Just one of these covers a significant part of the footprint of Manhattan," he said. The Instagram and WhatsApp owner on Wednesday raised the bottom end of its annual capital expenditures forecast by $2 billion, to $66 billion to $72 billion. It reported stronger-than-expected ad sales, boosted by AI-driven improvements to targeting and content delivery. Executives said those gains were helping offset rising infrastructure costs tied to its long-term AI push.

Meta to share AI infrastructure costs via $2 billion asset sale

(Reuters) -Meta Platforms is pressing ahead with efforts to bring in outside partners to help fund the massive infrastructure needed to power artificial intelligence, disclosing plans in a filing on Thursday to offload $2 billion in data center assets as part of that strategy. The strategy reflects a broader shift among tech giants -- long known for self-funding growth -- as they grapple with the soaring cost of building and powering data centers to support generative AI. The social media giant said earlier this week that it was exploring ways to work with financial partners to co-develop data centers to help finance its massive capital outlay for next year. "We're exploring ways to work with financial partners to co-develop data centers," Meta Chief Finance Officer Susan Li said on a post-earnings conference call on Wednesday. While the company still expects to fund much of its capital spending internally, some projects could attract "significant external financing" and offer more flexibility if infrastructure needs shift over time, Li said. The company did not have any finalized transactions to announce, she said. The disclosure in Meta's quarterly filing, however, signals that plans are firming up. In its quarterly filing on Thursday, Meta said it had approved a plan in June to dispose of certain data center assets and reclassified $2.04 billion worth of land and construction-in-progress as "held-for-sale". These assets were expected to be contributed to a third party within the next twelve months for co-developing data centers. Meta did not record a loss on the reclassification, which values the assets at the lower of their carrying amounts or fair value less costs to sell. As of June 30, total held-for-sale assets stood at $3.26 billion, according to the filing. Meta declined to comment for this story. CEO Mark Zuckerberg has laid out plans to invest hundreds of billions of dollars into constructing AI data center "superclusters" for superintelligence. "Just one of these covers a significant part of the footprint of Manhattan," he said. The Instagram and WhatsApp owner on Wednesday raised the bottom end of its annual capital expenditures forecast by $2 billion, to $66 billion to $72 billion. It reported stronger-than-expected ad sales, boosted by AI-driven improvements to targeting and content delivery. Executives said those gains were helping offset rising infrastructure costs tied to its long-term AI push. (Reporting by Echo Wang in New York; Editing by Sayantani Ghosh and Marguerita Choy)