📰 Key Highlights

An investigation by Japan’s Nikkei reveals that hidden off-balance-sheet debt at five US tech giants has surged eightfold in roughly four years to an estimated $1.65 trillion — exceeding the actual liabilities reported on these companies’ balance sheets and making it harder for investors to accurately assess their risk. The report points out that this type of hidden debt primarily comes from arrangements such as data center lease contracts and GPU supply agreements, and because it doesn’t show up directly on the balance sheet, transparency is lower. Take Meta as an example: its off-balance-sheet debt is estimated at around $420 billion, nearly triple its transparently disclosed debt. The report also mentions that Oracle faces a similar surge in hidden liabilities. As tech giants continue to ramp up AI-related capital expenditure, securing compute resources and data center capacity through non-traditional financing methods like leases and supply contracts, these off-balance-sheet liabilities are growing in scale and complexity — making it even harder for outsiders to grasp these companies’ true financial leverage and risk exposure. See the original link for full details.

💬 JudyAI Lab Perspective

A hidden debt surge of eightfold, approaching $1.65 trillion — that’s no small number. What Nikkei’s reporting uncovers is a structural funding problem in AI infrastructure, not just the old news of record-breaking capital expenditure.

The key thing we’ve observed is that arrangements like data center leases and GPU supply contracts are essentially financial engineering — securing compute power in ways that don’t hit the balance sheet, so reported liabilities look clean while actual leverage is far higher than the disclosed numbers suggest. Meta’s off-balance-sheet debt of around $420 billion is nearly triple its disclosed debt, and Oracle has the same problem. This reflects a trend: the funding pressure of the AI race has grown so intense that companies now have to rely on accounting tricks to spread it out. The traditional balance sheet alone can no longer sustain this pace of expansion. For AI builders, this is a reminder that whenever we evaluate any assumption about “AI infrastructure stability,” we should ask one more question: where’s the money actually coming from?

Next time you see a tech giant’s earnings report, don’t just look at the headline numbers — pay closer attention to the lease commitments and off-balance-sheet projects buried in the footnotes.

📅 Source Information

🔗 Related Reading