The AI Arms Race Joins Forces With the Literal Arms Race, Fueling $348 Trillion in Debt

Global debt had its biggest annual increase last year since the pandemic, according to a report by The Institute of International Finance.

Total debt now sits at a record $348 trillion, driven by global investment in national security and AI.

“Rising AI-related investment is emerging as a new driver of corporate borrowing and capital markets activity,” the report said.

The trend is very reflective of the year we’ve had. Fraying relationships (especially between the U.S. and its allies) and an increasingly tense geopolitical environment have led to increased investment in national security and defense initiatives. As defense becomes top of mind for world leaders, AI companies also made themselves a strong place in the category, as the technology entered the battlefield and AI sovereignty became the magic buzzword.

Meanwhile, big tech has been entrenched in an unprecedented AI infrastructure buildout to meet expected demand. AI spending is projected to reach $2.5 trillion globally this year, and an estimated $375 billion of that investment was made in 2025 alone.

The federal debt-to-GDP ratio is also projected to rise in the U.S. as Trump’s tariffs failed to generate enough revenue to reverse that trend, the report says.

“Evidence suggests much of the tariff burden has been absorbed by U.S. consumers and firms, weighing on private-sector balance sheets,” the report says. “Ahead of this year’s midterm elections, affordability concerns have already fed into Trump administration decisions to cancel or delay some previously announced tariffs on agricultural products and furniture imports.”

The good news for Silicon Valley, though, is that current financial conditions should continue to make borrowing easier for national priorities like defense and AI.

“A powerful new wave of global capital expenditure ‘supercycles’ is set to reinforce this momentum, with large-scale investment in AI-driven data centers, energy security and transition, and resilient infrastructure emerging as a major growth engine for global debt markets,” the report says.

But that global spending boom comes with risks.

“This expansion is unfolding at a time when global growth remains broadly resilient, raising questions
about whether the combined force of fiscal, monetary, and regulatory stimulus—together with capex-driven private borrowing—could eventually result in episodic overheating and stretched valuations in some areas,” the researchers say.

TLDR: an AI bubble is still very much on the table.

Tech giants in the AI space have been engaged in an infinitely expanding and tangled web of multibillion-dollar dealmaking amongst each other that many have deemed to be circular.

But as the AI spending commitments of these companies reach once-unimaginable highs, some investors are increasingly worried that this spending is being secured perhaps a bit too much by debt.

Bonds issued by American AI companies are on track to reach yet another record high in 2026, according to Wednesday’s report. In fact, AI-related funding needs are expected to be the biggest driver for U.S. corporate bond issuance in 2026, according to a Barclays report. Oracle, one of the major financiers of the so-called AI revolution, has taken on more than $100 billion in debt to secure its investments.