Analysts see risks for the financial system in the AI boom - ForkLog

# Analysts See Risks to the Financial System in the AI Boom

The investment boom around artificial intelligence, which supported the global economy in 2025, is itself becoming a source of macro-financial risks. This is stated in the annual report of the Bank for International Settlements (BIS).

The document discusses debt financing for AI infrastructure, inflated company valuations, the growth of private lending, and opaque deals between hyperscalers, chipmakers, and laboratories.

AI helped growth, but increased vulnerabilities

According to the BIS, in 2025 the global economy showed resilience to tariff and geopolitical shocks. One factor was optimism surrounding artificial intelligence: it supported capital expenditures, trade in intermediate goods, and easy financial conditions.

However, in 2026 the set of risks expanded. The bank’s analysts identified four pressure points:

• the threat of persistent inflation;
• concerns about the sustainability of AI investments;
• rising financial vulnerabilities;
• a deterioration in fiscal positions.

“Optimism about AI may not last, despite the promise of future productivity growth,” the report says.

According to the BIS, the current rise in capex may prove unsustainable if production hits bottlenecks. Among them, the bank cited electricity, advanced semiconductors, and networking equipment.

Capex is outpacing cash flow

In the report, capital expenditures for AI by the five largest hyperscalers in 2025–2026 are estimated at more than $1 trillion (a combined figure for the two years). According to the BIS, these commitments are already outstripping the profits and free cash flow of some companies, forcing them to raise debt financing.

Growth of hyperscaler AI capex and comparison of the current cycle with past investment booms. Source: BIS Annual Economic Report 2026.

“Disappointment in returns could trigger a sudden withdrawal of financing and turn the capex boom into a protracted investment slump with potential consequences for financial conditions,” analysts warned.

The bank compared the current cycle with previous episodes of technology overheating: canal mania of the 1830s, Britain’s railway mania of the 1840s, the electrification euphoria of the late 1920s, and the late-1990s dot-com boom. BIS called a common feature of these episodes the real technological breakthrough that attracted more capital than commercial results could later justify.

Circular financing became a separate risk

In the view of the report’s authors, credit markets, private lending, and contractors building data centers, power and related infrastructure may be vulnerable. If hyperscalers slow down or stop aggressive capex, borrowers across the supply chain could lose the revenue needed to service their debt.

Another vulnerability is opaque private deals within the artificial intelligence sector. This refers to circular financing, in which hyperscalers or chip manufacturers receive stakes in AI labs and cloud providers, while the latter take on multi-year commitments to procure chips or computing resources.

“The terms of such deals are usually poorly disclosed, creating a risk of rehypothecation of the same asset,” the document says.

The BIS also reminded that a correction in the United States could spread rapidly around the world. American stocks account for about 64% of the MSCI Global index, so a reassessment of AI companies could affect household wealth, consumption, and global financial conditions.

The memory market showed the scale of demand

Separately, the memory chip market became an additional indicator of strain in the AI supply chain. On June 24, U.S. semiconductor manufacturer Micron reported record revenue for the third quarter of fiscal 2026 — $41.46 billion. In prepared remarks for the investor call, the company said it entered into 16 strategic customer agreements. Such contracts are usually for five years — from 2026 through the end of 2030; automotive agreements typically have a three-year term.

According to Micron, for already signed agreements, including those signed after the end of the quarter, RPO is about $100 billion. The company emphasized that the figure does not equal all expected future revenue. The manufacturer also expects $22 billion in customer deposits and related financial obligations under already signed agreements. Of these, about $18 billion will be in cash deposits.

“Right now, we don’t have clarity on when memory [chips] supply will be able to catch up with the growing demand,” said Micron CEO Sanjay Mehrotra.

In the reported quarter, DRAM brought Micron record $31.3 billion, or 76% of total revenue. The company linked the growth to tight industry conditions, a favorable sales mix, and rising prices.

At the same time, the increase in chip prices has already become the subject of a legal dispute. On June 25, a class-action antitrust lawsuit was filed in the Northern District of California against Samsung Electronics, Samsung Semiconductor, SK Hynix, SK Hynix America, and Micron Technology.

According to MLex, the plaintiffs believe that the three largest DRAM producers coordinated supply constraints and price increases. The claims also mention shifting priorities toward more expensive AI products, including HBM memory.

Earlier, it became known that growth in demand for AI infrastructure has become a problem for tech giants due to spending on memory, electricity, and data centers.

Recall that in November 2025, the media reported on Micron’s plans to invest $9.6 billion in AI memory chip production in Japan.