Sources: Meta is looking to raise $3B in equity and $26B in debt, from private capital firms, including Apollo Global and KKR, to fund its data center build out
www.ft.com/content/aff1... @ivanthek : WTF? — This is not exactly a vote of confidence in their own strategy. — www.ft.com/content/aff1... [image] X: Luke Kawa / @ljkawa : Private credit and AI, match made in this bull market's heaven. $META Alexander Doria / @dorialexander : I'm afraid AI is just becoming a way to rebrand unsexy business move: pivoting from markets with little growth prospect, overdue infra updates, downsizing teams... @zephyr_z9 : This is not good Meta has a huge balance sheet Xai is issuing debt at 12.5% interest We may be in bubble territory Jason Kint / @jason_kint : I find this absolutely nuts considering the claims and evidence in the Facebook trial starting in just a few weeks in Delaware court. [image]
Context & Ripple Effects
Meta had already signaled that AI infrastructure would require materially higher spending when it lifted its 2024 capex outlook amid investor concern over returns. Its strong 2025 Q1 results supplied operating cash flow, but this report shows the build-out was large enough to draw on outside capital as well as Meta’s balance sheet.
The proposed raise is an early step in the financing arc later associated with Louisiana data-center structures: subsequent coverage examined a $27B Louisiana financing arrangement designed to keep construction debt off Meta’s balance sheet. That makes the Apollo and KKR discussions significant as an example of private capital becoming part of hyperscale AI infrastructure funding.
First-order effects
- Meta would obtain a reported $29B split between $3B of equity and $26B of debt for data-center construction, while Apollo, KKR, and other private-capital providers would take on financing exposure to the project.
- The funding route expands Meta’s available construction capital without relying solely on operating cash flow, following its earlier increase in AI-infrastructure spending plans.
Second-order effects
- Private-credit and infrastructure investors gain a larger role in funding AI capacity, potentially making project-finance terms, collateral, and ownership structures more consequential to data-center expansion.
- Other large AI investors face a clearer precedent for combining equity and private debt to fund compute build-outs, rather than treating infrastructure spending only as conventional corporate capex.
Third-order effects
- If replicated, AI data centers could increasingly be financed as long-lived infrastructure assets, separating the economics and risks of physical compute capacity from the technology companies that use it.
- That shift may concentrate the ability to build frontier-scale capacity among companies that can attract both operating cash flow and large private-capital commitments; the durability of that model will depend on whether utilization and AI-derived revenue support the financing.
The trend: AI infrastructure finance is evolving from balance-sheet-funded capex toward private-capital-backed, project-style financing for increasingly expensive compute assets.