Meta Faces Higher Costs on $12 Billion AI Data-Center Financing

Meta and BlackRock are jointly investing $14 billion in a gigawatt-scale data center designed to meet surging demand for artificial intelligence and cloud computing capacity. Such projects require heavy upfront spending, long construction periods and substantial debt financing. Adequate coverage for property damage, equipment failures and business interruption is therefore critical to protecting lenders whose capital remains exposed throughout development and operation.

The project now faces an insurance shortfall because carriers are struggling to absorb the cost of providing full coverage, leaving billions of dollars of risk underinsured, according to the latest report. If a major loss occurs, insurance proceeds may fall short of the amount needed to cover damages or repay financing. The gap raises the financial exposure of participating lenders and highlights a growing constraint on funding ever-larger AI infrastructure projects.

Meta’s partnership with BlackRock highlights how technology companies are tapping outside capital to fund the rapidly rising cost of artificial-intelligence infrastructure. By using a joint-venture and project-finance structure, Meta can expand computing capacity while limiting the burden on its own capital spending and free cash flow, as competition for the data centers and power needed to train and run AI models intensifies.

The companies have formed a venture valued at about $14 billion to develop a 1-gigawatt AI data center campus in El Paso, Texas. The facility is expected to begin coming online in 2028, adding a major block of computing capacity for Meta’s AI expansion. The financing structure allows the partners to share development costs and risks while accelerating construction of the power-intensive infrastructure.

The generative AI race is driving capital spending by major technology companies to new highs. Meta Platforms plans to invest as much as $145 billion in 2026 to expand its data centers, servers and chip supplies. Raising the money through a new share issuance could ease borrowing pressure but may dilute existing shareholders’ stakes and earnings per share.

Meta is reportedly evaluating the issuance of tens of billions of dollars in new shares to supplement funding for its 2026 AI infrastructure budget, though the size and timing have yet to be finalized. The report emerged shortly after Alphabet completed about $85 billion in financing, underscoring how technology giants including Meta and Google are expanding external fundraising as they compete for AI computing capacity.

Meta is accelerating construction of data centers and purchases of servers and computing chips needed to support generative AI, with its vast infrastructure requirements increasing funding pressure. Raising money in the bond market allows the company to finance its expansion while preserving cash flexibility, but has also sharpened investor scrutiny of whether its AI services can generate advertising and product revenue.

Meta recently completed a $25 billion bond offering to fund AI infrastructure in 2026 and raised the upper end of its full-year capital spending plan to $145 billion. Despite strong demand for the bonds, investors continue to question the returns on Meta’s heavy AI investment because the company has yet to provide a clear timeline for monetizing its products.