On August 16, 2026, the Wall Street Journal reported that nine major technology companies, including Alphabet Inc. (NASDAQ:GOOGL) and Meta Platforms, Inc. (NASDAQ:META), carry roughly $3 trillion in AI-related commitments that do not appear on their balance sheets. That figure is about five times the roughly $600 billion in capital expenditures the companies reported over the past year and about triple their combined on-balance-sheet lease and debt liabilities, the Journal found. The companies show $248 billion in leases on their balance sheets but carry roughly $900 billion to $1.2 trillion more off of them and disclose $356 billion in on-balance-sheet debt, separate from roughly $1.9 trillion in additional purchase commitments the Journal identified.
Alphabet Inc. (NASDAQ:GOOGL) alone has accumulated an estimated $811 billion in purchase commitments, up sharply from $322 billion three months earlier, according to the Journal’s data.
Bull Case
Off-balance-sheet financing is legitimate. It is a widely used accounting tool, not evidence of trouble on its own. Structuring long-term data center leases and chip purchase agreements this way is standard practice under current accounting rules. It lets companies like Alphabet and Meta Platforms, Inc. (NASDAQ:META) commit to AI infrastructure without immediately inflating reported debt, preserving financial flexibility and credit ratings while capacity gets built.
The scale of these commitments shows contracted demand rather than speculative spending. Optimists see soaring demand for AI tools, which has already produced hardware shortages, as evidence that revenue to cover these obligations will materialize over time. Companies do not typically sign multi-year, multi-billion-dollar purchase and lease commitments without confidence in the underlying demand driving them.
Alphabet Inc. (NASDAQ:GOOGL) specifically has a track record that supports this kind of aggressive capital commitment. Alphabet’s own reported return on prior technology investment has historically been strong. As one of the most cash-generative companies in the index, it has the operating cash flow to support large purchase commitments without the same balance-sheet strain a smaller or less profitable company would face.
Bear Case
The obligations are growing faster than the reported spending investors already track. The Journal found off-balance-sheet AI commitments are expanding faster than traditional capex. It means the headline capital-spending figures companies report each quarter increasingly understate the real financial exposure shareholders are underwriting.
Credit markets are already pricing in more risk. The cost of insuring major technology companies’ debt against default has climbed to record highs as AI infrastructure spending strains balance sheets with heavy investment-grade bond issuance, up 36% year over year. It pushes longer-term borrowing costs higher across the sector.
Meta Platforms, Inc. (NASDAQ:META)’s financing structure has drawn specific scrutiny as a sign of how aggressive this has become. Meta has used single-tenant, off-book financing vehicles for major data-center projects, a structure commentators have flagged as characteristic of late-cycle credit engineering, meaning the debt exists and must eventually be serviced even though it does not show up on Meta’s own balance sheet today.
Conclusion
This is a story about financial structure catching up to spending ambition, not a story about AI demand itself failing. The bull case rests on off-balance-sheet financing being a normal tool that lets cash-rich companies like Alphabet Inc. (NASDAQ:GOOGL) fund AI buildout without immediately straining their credit profile, backed by demand strong enough to justify the commitments. The bear case rests on the size and growth rate of these hidden obligations, now roughly triple visible debt and leases, and on early signs that credit markets are starting to price the risk that not every dollar of AI spending converts into revenue on schedule.
READ NEXT: Warren Buffett “Blew It” on Alphabet (GOOGL) And Made It Berkshire’s Third-Biggest Bet and Sony Group (SONY) and Taiwan Semiconductor (TSM) Are Betting $4.7 Billion on the “Eyes” of AI Machines.
Disclosure: None. This article is originally published at Insider Monkey.
Hedge Fund Resource Center