Meta seeks $29b for US-based AI data centers

Meta Platforms is reportedly seeking to raise US$29 billion from private capital firms to build AI data centers in the US, according to the Financial Times.

The plan includes US$3 billion in equity and US$26 billion in debt.

Meta is in advanced talks with Apollo, KKR, Brookfield, Carlyle, and PIMCO.

The company is working with Morgan Stanley on the financing and may pursue more capital or make the debt more tradeable.

This effort follows Meta’s increasing AI investments, including US$14.8 billion in Scale AI.

CEO Mark Zuckerberg has said the company could allocate up to US$65 billion in 2025 for AI infrastructure.

🔗 Source: Reuters

🧠 Food for thought

1️⃣ Meta’s financing pivot reflects a new era in big tech capital strategy

Meta’s pursuit of $29 billion in private capital funding marks a dramatic shift from its historical approach to financing growth.

In its early days, Facebook raised a total of just $1.3 billion before going public, starting with Peter Thiel’s $500,000 investment in 2004 and culminating with Goldman Sachs’ $500 million investment in 2011 when the company was valued at $50 billion 1.

This external financing approach, seeking $3 billion in equity and $26 billion in debt, suggests that even cash-rich tech giants are reconsidering how to fund the massive capital expenditures required for AI infrastructure without depleting their balance sheets.

The scale reflects the extraordinary cost of AI infrastructure compared to previous technological transitions, with Meta’s $29 billion fundraising representing more than 22 times the company’s entire pre-IPO funding.

This financing strategy mirrors the infrastructure funding models traditionally used by utilities and telecommunications companies rather than software firms, potentially signaling a new phase in how tech companies finance critical infrastructure.

2️⃣ The AI data center boom faces unprecedented energy and capacity challenges

Meta’s massive investment comes amid projections that global data center capacity demand will increase by 19-22% annually through 2030, with AI workloads expected to account for 70% of all data center capacity by decade’s end 2.