The International Monetary Fund has flagged the possibility that financial markets could face instability if the pace of AI-related investment, corporate debt accumulation, and profit expectations outstrips the technology's actual gains in productivity.
Speaking at the gathering, IMF Managing Director Kristalina Georgieva pointed to measurable output gains tied to the spread of artificial intelligence across a number of economies, with the United States and Asia's hardware-producing nations prominent among them, while noting that the same technological momentum was also allowing financial vulnerabilities to build up alongside whatever benefits it brought.
Writing in a Thursday article, Georgieva put forward the view that capital flowing into artificial intelligence infrastructure might one day surpass the vast outlays that once built out railway networks, electrical grids, and the internet. Outlays on data centres, computing infrastructure, and associated technologies have fed fresh demand across a range of economies, with early returns pointing in a favourable direction.
Georgieva noted that the financial frameworks underpinning that growth phase carry exposure to a potential reversal should assumptions about AI's earnings potential turn out to be inflated. How technology firms are funding their investments has drawn attention, with a notable shift toward borrowing from debt markets rather than other sources of capital.
Georgieva noted that large technology firms funding AI infrastructure have come to represent roughly a tenth of all investment-grade bond issuance — a share that barely registered at under one per cent just two years prior. Her remarks drew attention to how few players dominate the sector and to the web of overlapping ties forming between technology firms and those building artificial intelligence systems.
Money leaving large technology companies finds its way to AI research organisations, which in turn spend a share of those received funds back with those same corporations on computing infrastructure and hardware. Where projected revenues fall short of expectations, these structures carry the potential to deepen financial pressure.
The IMF's assessment noted that weaker-than-expected revenue or profit figures from AI developers could weigh on investor sentiment and put downward pressure on technology sector valuations. Breaches or malfunctions tied to AI systems of growing capability carry the potential to introduce fresh volatility into an already unsettled landscape.
Pricing strains across a debt-laden, densely interconnected AI sector could spread beyond technology under deteriorating market conditions, pulling banks, investment funds, and a range of other financial actors into the disruption. Where AI-linked financial holdings are distributed across multiple countries, a downturn taking root in one market carries the potential to ripple outward beyond that nation's borders.
Georgieva pointed to artificial intelligence as a technology carrying the potential for substantial gains in economic output over time. Projections put the uplift to worldwide output at roughly half a percentage point per year across the next ten years, with the broader economy standing to benefit from the technology's reach.
Georgieva noted that realising those benefits would depend on organisations fundamentally rethinking how their products, processes, and working methods are structured, rather than merely layering new technology on top of existing frameworks.
Employment-related risks were among the areas the IMF flagged in its assessment. Findings from the research place roughly 40 per cent of jobs globally within the reach of artificial intelligence's influence, a share that climbs to around 60 per cent when the focus narrows to wealthier, more developed economies.
Exposure to that influence does not, on its own, guarantee that roles will be eliminated. Across the workforce, AI tools are reshaping the landscape unevenly — lifting output in certain roles while narrowing the demand for others or altering the skills those positions require. No definitive pattern of widespread job losses had yet materialised, the IMF noted, though the organisation pointed to shifts already under way in how people carry out their work.
Georgieva urged governments and financial regulators to put stronger safeguards in place against vulnerabilities that could materialise in the system. Tighter oversight of financial institutions, closer scrutiny of concentrated investment positions, and stronger defences against cyberattacks were among the measures she put forward.
Gauging how artificial intelligence reshapes price levels, worker output, and broader economic activity was identified as a responsibility falling within the remit of monetary policymakers. Fiscal authorities across member states received an IMF advisory to prepare for potential shifts in tax collection, earnings from labour, and distributional outcomes as the technology spreads into broader use.
Spending on data centres, advanced computing hardware, and the power networks required to run them has climbed across the technology sector, with artificial intelligence drawing much of that capital into the United States.
