The International Monetary Fund has warned that the global economy has become too dependent on artificial intelligence investment, raising the risk that a sharp reversal in technology spending could trigger financial instability and weaken economic growth.
IMF Managing Director Kristalina Georgieva said Monday that the world economy was relying heavily on AI-related investment to sustain growth at a time when energy costs and geopolitical tensions were weighing on other parts of the global economy.
“We have a world economy that is standing on one leg on the AI leg. If this leg wobbles, that can lead to financial stability troubles,” Georgieva told CNBC’s JP Ong on the sidelines of the Association of Southeast Asian Nations summit in the Philippines.
Register for the next Tekedia Mini-MBA.
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
She warned that a sudden loss of investment confidence and withdrawal of capital from the AI sector could have consequences beyond technology companies, given the industry’s growing influence on trade, corporate investment and financial markets.
One of the IMF’s concerns is the expansion of circular financing arrangements among AI companies. Under these structures, investors provide capital to AI businesses that subsequently use some of the money to purchase chips, cloud computing capacity or other services from companies connected to the original investors.
The IMF estimates that these arrangements have grown to slightly more than $1 trillion, with about two-thirds financed through debt. Georgieva said the total was not yet large enough to threaten the global economy on its own, but the speed of its expansion warranted attention.
“It’s not a huge size, but it is a rapid increase,” she said.
The concern is that financing and commercial demand within the AI ecosystem may be more closely connected than headline investment figures suggest. When companies invest in customers, suppliers or infrastructure partners that later generate revenue through those relationships, capital can circulate within the same network. This can support rapid expansion, but it also makes the system more vulnerable if financing conditions deteriorate or expected demand fails to materialize.
Nvidia Sits at the Center of Concern
Nvidia sits at the center of the AI infrastructure buildout, supplying advanced processors while also providing financial support to parts of the ecosystem purchasing computing capacity. That position has made the chipmaker important not only as a technology supplier but also as a participant in the financing arrangements supporting the industry’s expansion.
The company’s recent turmoil has added another source of uncertainty. More than $1 trillion has been wiped from Nvidia’s market value over the past month, while reports surrounding former chief executive Jensen Huang’s extramarital affairs, including an alleged relationship with a woman suspected of links to Chinese intelligence, have drawn attention to corporate governance and the company’s exposure to political and regulatory risks.
The resulting sell-off in AI-linked stocks has intensified scrutiny of Nvidia’s influence over the sector. Investors are assessing how a change in leadership, or a reduction in the company’s willingness to finance customers and infrastructure projects, could affect the wider investment cycle.
UBS analyst Tim Arcuri estimated in a recent note that a pullback in Nvidia financing could remove $50 billion or more from the system. Such a reduction could affect companies that rely on Nvidia-backed investments or commercial commitments to expand data centers and secure computing capacity.
The potential impact is amplified by borrowing costs. Higher interest rates have made debt financing more expensive, increasing the financial burden on companies that are expanding infrastructure ahead of future revenue. If Nvidia or other major technology companies reduce their financial support, some projects could face difficulty replacing that capital on comparable terms.
This has resulted in a vulnerability that extends beyond Nvidia’s share price. Data center operators, cloud providers and AI developers have made large commitments based on expectations of sustained demand for computing power. A financing slowdown could force companies to delay projects, reduce spending or renegotiate commercial arrangements.
The risk is not that all AI investment will disappear, but that capital spending could slow faster than businesses have prepared for. Companies with heavy debt obligations and large infrastructure commitments would be particularly exposed if expected customer demand or financing support weakened.
AI Investment Buoys a Growing Share of Global Growth
Georgieva’s warning comes as AI-related spending has become an important driver of global economic activity. Investment in semiconductors, data centers and computing infrastructure has supported manufacturing, trade and capital expenditure, particularly in major economies.
The IMF raised its 2026 global growth forecast to 3.1% in October, as AI investment helped offset economic pressures, including the energy shock associated with the Iran war.
The scale of the technology’s economic influence is also visible in international trade. Capital Economics estimated that as much as 80% of the recent expansion in global exports was concentrated in AI-related goods. These include semiconductors and other equipment required to build and operate AI infrastructure.
The IMF estimates that approximately 40% of the world’s workforce will be affected by AI, indicating that the technology’s influence extends well beyond the companies developing models and manufacturing processors. However, the immediate contribution to economic growth is being driven heavily by investment in physical infrastructure and computing capacity.
Georgieva said investor expectations were running ahead of what AI had demonstrated so far. As valuations rise, the gap between anticipated future returns and proven commercial performance becomes more important.
“We tend to rush ahead,” she said, adding that “some corrections” were necessary.
The IMF had already warned in October that a short-term disappointment in AI could trigger a disorderly correction across financial markets. In its worst-case scenario, if the technology failed to meet elevated expectations, US equity prices could fall 20% in 2026, while global economic output could end up 1.1% below the institution’s baseline projection by 2028.
That scenario illustrates how a technology-sector correction could spread through broader financial markets. A sharp fall in technology valuations could weaken investor wealth, restrict access to capital and reduce corporate investment. The consequences would extend to economies that have benefited from semiconductor exports and the construction of AI infrastructure.
Global equities, including US chipmakers, have already experienced a difficult period, putting the AI investment thesis under pressure ahead of Nvidia’s earnings and Anthropic’s planned blockbuster stock market listing.
Those events will test investor appetite for companies whose valuations depend heavily on expectations of future AI growth. Nvidia’s results will provide an important indication of chip demand and spending across the industry, while Anthropic’s listing will offer another measure of how much investors are willing to pay for exposure to AI development.
Georgieva described the recent market volatility as orderly and said it was unlikely to develop into a crisis capable of undermining the global economy. Her assessment, however, came with a warning that investors and policymakers should remain alert.
“We have to be watchful,” she said.
The key issue for the global economy is the concentration of growth around an industry whose long-term commercial potential remains difficult to measure. AI investment has delivered substantial demand for chips, computing infrastructure and related equipment, but the financing structure supporting that expansion creates exposure if expectations change.
A correction would not necessarily mean the technology has failed. However, analysts say it could instead expose companies that have borrowed heavily, relied on connected investors, or committed capital ahead of sustainable customer demand.



