News · Meta and Blue Owl split ownership of the $27B Hyperion campus 80/20, keeping most of the build off Meta's balance sheet

Oct, 214 min to read
Frontend

Meta and Blue Owl split ownership of the $27B Hyperion campus 80/20, keeping most of the build off Meta's balance sheet

A joint venture in Richland Parish, Louisiana shows how Meta is financing AI capacity through leases and guarantees rather than direct ownership.

Who owns Hyperion, and who pays for it

On October 21, 2025, Meta and funds managed by Blue Owl Capital formed a joint venture to develop and own the Hyperion data center campus in Richland Parish, Louisiana. The split is lopsided: Blue Owl's funds hold 80% of the venture, Meta keeps 20%.

The two sides committed to fund their pro rata shares of roughly $27 billion in total development costs — buildings plus what the release calls the long-lived power, cooling, and connectivity infrastructure. Meta contributed land and construction-in-progress assets that had previously been classified as held-for-sale, Blue Owl's funds put in about $7 billion in cash, and Meta received a one-time $3 billion distribution back out of the venture.

So Meta already has thousands of construction workers on site — its own account — while structuring the ownership so that an outside investor group carries the majority stake and the bulk of the capital.

Meta builds it, then leases it back

Meta isn't stepping away from Hyperion; it's providing construction management and property management services, and it signed operating leases for all of the campus facilities once construction finishes. Those leases run a four-year initial term with options to extend.

A short base term with renewals gives Meta room to walk if its capacity needs change. To make that optionality cheaper, Meta also granted the venture a residual value guarantee covering the first 16 years of operations: if a lease isn't renewed or is terminated under certain conditions, Meta would make a capped cash payment based on the then-current value of the campus.

That guarantee is the tell. Meta gets the flexibility of a tenant while absorbing the downside risk of an owner. The asset sits in a venture Meta only partly owns, but the economic exposure stays close to home.

The debt layer behind the equity

Blue Owl's contribution isn't all equity. The release states that a portion of the capital it raised will be funded by debt issued to PIMCO and select other bond investors through a private securities offering. Morgan Stanley served as Meta's exclusive financial advisor and as sole bookrunner on that offering.

That stacks three sources of money against one campus: Blue Owl fund equity, bondholder debt, and Meta's own contribution and lease payments. Fixed-income investors like PIMCO get a long-dated, infrastructure-backed instrument; Meta gets capacity without booking the full $27 billion as its own capital expenditure.

What the financing tells you about the AI capacity you'll actually use

CFO Susan Li framed the deal in terms of delivery: her stated point is that Meta's AI ambitions are realized through its ability to build the infrastructure to support them. Hyperion is that infrastructure — the compute that will eventually sit behind recommendation, generation, and assistant features across Instagram, WhatsApp, and Messenger.

Our AI ambitions will be realized through our ability to deliver the infrastructure to support it.Montana Labs

The frontend implication is quieter than the headline number. The models and features users touch are increasingly downstream of financing engineering — 80/20 ownership splits, four-year leases, residual value guarantees, and private bond offerings. When a company structures $27 billion this carefully to keep exit optionality, it's a signal that even a hyperscaler is hedging how much of this AI capacity it wants to be locked into owning outright.

Find this story relevant to you?

Contact us to find a unique solution

Contact us

Need an AI engineering partner that can actually build?

We help businesses integrate AI, build AI-powered products, automate high-value workflows, and modernize the software systems behind them.

Get in touch

Related reading

More analysis around product delivery, operational AI, and the systems work that makes deployment hold up in reality.

Jul, 134 min to read
Frontend

DNP put ChatGPT Enterprise in front of ten departments and treated the chat window as the interface

Jul, 134 min to read
Frontend

AdventHealth deploys ChatGPT across nine states by treating adoption as the product

Jul, 134 min to read
Frontend

AP+ uses Codex to build behaving payment prototypes, not just clickable screens