Hidden liabilities tied to artificial intelligence spending have climbed sharply at five major U.S. technology companies, reaching an estimated $1.65 trillion, according to a Nikkei study. The report says those obligations expanded roughly eightfold in about four years as the race to build AI capacity accelerated.
The increase is linked to commitments that may not be obvious in headline debt figures, including large data center leases and long-term GPU supply contracts. These arrangements can support rapid AI expansion, but they also add significant financial obligations outside the traditional view of corporate borrowing.
The study highlights companies including Meta and Oracle, showing how the push to secure computing power and infrastructure is reshaping balance-sheet risk across the sector. As demand for AI services grows, tech groups appear willing to lock in major capacity commitments to avoid supply shortages and keep up with competitors.
The findings suggest that the cost of the AI boom is extending well beyond capital spending alone. For investors and analysts, the rise in opaque AI funding may bring renewed attention to how future liabilities are disclosed and how much financial strain could build as companies continue scaling their AI operations.
