IMF Chief Georgieva Warns AI Could Lift Growth While Deepening Global Risks

Table of Contents

• You might want to know
• AI’s promise and the pressures facing the global economy
• Key Insights Table
• Afterwards...

You might want to know

• How could artificial intelligence boost global growth while widening economic inequality?
• Why does the IMF chief see public debt, energy prices, and AI investment as connected risks?

AI’s promise and the pressures facing the global economy

SINGAPORE — Artificial intelligence is increasingly viewed by investors and governments as a source of productivity and economic growth. Yet the same investment surge is creating new pressures for governments and financial markets. International Monetary Fund Managing Director Kristalina Georgieva warned that AI is arriving at a time when energy costs are rising and public debt is already high. She urged policymakers not to postpone difficult decisions about debt and fiscal policy.

Speaking at a Wednesday event in Singapore, Georgieva described AI as a force that is becoming central to countries’ relative prospects in the world economy. “Love it, hate it, or fear it, AI is here,” she said. In her view, the technology could influence which countries gain economic ground, but its effects will depend on how broadly the benefits are shared and how well governments manage the risks.

Her remarks came ahead of a series of annual meetings of the IMF and World Bank that kick off next week. Georgieva portrayed the global outlook as being pulled in opposing directions. The war in the Gulf, now in its eighth month, is contributing to what she called a “negative energy supply shock.” At the same time, the surge in AI investment is generating a “positive demand shock.” The two forces are interacting with an already “underwhelming” growth outlook for this decade, and their combined impact is “highly uneven across the world.”

The scale of AI investment helps explain why it is attracting such attention. Georgieva said global spending on AI, as a share of GDP, will reach and likely exceed the amounts invested in building railroads, electricity grids, and telecommunications networks. AI hardware and related technology products already make up more than a tenth of world goods trade, according to her remarks. The IMF estimates that AI could add up to half a percentage point to annual world growth if the technology is adopted effectively.

Georgieva illustrated the potential scale of that contribution with a comparison: “Going from 3% to 3.5% over a decade — that is like adding an economy the size of ASEAN to the world economy.” The comparison underlines why AI is often presented as a possible answer to weak productivity and sluggish growth. If new tools help workers and businesses produce more with existing resources, the resulting gains could accumulate across industries and countries.

But the benefits are unlikely to be evenly distributed. Countries that are less integrated into the global AI supply chain may see less investment and fewer direct gains. Georgieva warned that the boom could therefore increase the risk of widening economic inequality around the world. Advanced technology can create new opportunities, but countries need access to infrastructure, skills, investment, and relevant supply chains to capture them. Without those conditions, the expansion of AI could reinforce differences in economic performance rather than reduce them.

The investment boom also adds to concerns about inflation. Georgieva said that “the AI building boom is inflationary,” alongside energy and food shocks, tariffs, and defense spending. A rapid expansion in the construction of data centers and related infrastructure can raise demand for materials, equipment, electricity, and financing. At the same time, supply disruptions and higher costs for essential goods can add to price pressures that central banks are already trying to contain.

Energy costs are a particular concern. Oil prices have remained above $100 per barrel as the conflict in the Middle East continued, with few signs of a diplomatic off-ramp. Retail diesel prices also rose to record highs as constrained refining capacity squeezed energy supplies. These pressures can feed into the cost of transportation, production, and household spending, while making it more expensive to build and operate energy-intensive technology infrastructure.

Higher inflation expectations and borrowing needs are also being reflected in bond markets. Yields in the United States, Germany, and Japan have surged to their highest levels in decades. Companies connected to AI are issuing growing amounts of long-term private debt, adding to competition with governments for available capital. Georgieva noted that at least part of the rise in yields may also reflect expectations of faster economic growth. Even so, more expensive borrowing can weigh on both public budgets and private investment.

Public debt is another central part of the challenge. Georgieva said global public debt is close to its highest level since World War II and is on track to soon exceed 100% of GDP. Advanced economies, she said, are the “worst offenders.” For 17 years, governments had “a relatively easy ride” because interest rates remained below growth rates. That situation has now changed: “Higher interest rates now put an end to that.”

The relationship between interest costs and economic growth matters because it shapes how easily governments can stabilize their debt. Georgieva said the interest-to-growth differential is now “much less favourable” and “set to climb higher.” As a result, relying on growth alone to lower debt ratios, without fiscal action, is “out of reach in the near term.” Governments may need to make choices about spending, revenue, and priorities even as they face pressures to invest in energy security, defense, and technology.

Signs of fiscal strain are already visible in Europe. Spreads over German bunds are widening not only for France and Italy, but also for Ireland, Portugal, and other countries that reduced debt and deficits after the euro-area crisis. Georgieva linked this pressure to a series of shocks that have increased public debt. She also noted that most countries’ fiscal deficits remain above pre-pandemic averages. In her words, “fiscal space is crying out for replenishment.” The point is not simply that governments should cut spending, but that they need enough room in their budgets to respond when future crises arise.

AI itself could also pose a financial stability risk. Strong corporate earnings have supported share prices and created wealth effects, Georgieva said. However, the market outlook could change if companies fail to meet expectations. She warned that hyperscaler leverage, combined with large and growing global holdings of U.S. equities, could turn disappointing earnings into a far-reaching shock. In other words, optimism about a fast-growing technology sector may become a source of broader vulnerability if valuations and borrowing depend on results that do not materialize.

Georgieva referred to Amara’s Law, which holds that people tend to overestimate a new technology’s effects in the short run and underestimate them in the long run. She said the economy is in a transition between today’s AI building boom and the future arrival of AI’s wider benefits. It is during that interval, she argued, that the world may pass through a “period of maximum risk.” Investment and expectations can surge before productivity gains are fully established, leaving markets exposed if the transition proves slower or more uneven than anticipated.

Her proposed first line of defense is regulation and supervision. Rules and oversight can help authorities monitor leverage, assess risks, and improve resilience as AI-related investment expands. Georgieva also said, “Now may be a good time for a prudently hawkish bias in many countries’ monetary policy.” That comment reflects the challenge facing policymakers: they must guard against persistent inflation without unnecessarily undermining economic activity or the investment that could support longer-term growth.

Taken together, Georgieva’s assessment is neither a rejection of AI nor an unqualified endorsement of the current boom. AI could make a meaningful contribution to global growth, but the result depends on implementation and on whether gains reach economies beyond the main technology supply chains. Meanwhile, rising energy costs, high interest rates, public debt, and financial-market exposure can limit governments’ ability to respond if conditions worsen. The central policy task is to encourage productive innovation while preparing for the possibility that expectations, costs, or markets may shift quickly.

Key Insights Table

Afterwards...

The coming years will test whether AI’s investment surge can translate into lasting productivity gains without deepening financial and economic divides. Policymakers will need to monitor inflation, debt, market leverage, and the distribution of technology’s benefits at the same time. The outlook is not predetermined: sound oversight, responsible fiscal choices, and broader access to AI capabilities may help turn the current boom into durable growth while limiting the risks Georgieva identified.