The Stabilisation Gap: Artificial Intelligence Exposure and Macroeconomic Policy Space in the World's Ten Largest Economies

Economic stability is the art of keeping a nation's output and prices steady while the world around it shifts. When a new technology arrives, it changes the speed at which an economy naturally grows, a speed economists call the trend. To keep the economy on track, central banks and governments must adjust their tools: interest rates to manage spending, and taxes or borrowing to manage debt. For decades, the leading theory on how to do this has relied on a specific mathematical relationship between how much people save for the future and how fast the economy grows. If a powerful new technology arrives, it does not just create jobs or products; it changes the fundamental math of how interest rates and debt interact. The question facing the world's largest economies is whether they are ready for the specific kind of shock that artificial intelligence brings, and whether their current financial tools are strong enough to handle it.

A researcher at the University of Sevilla has examined this question by looking at the ten largest economies in the world: the United States, China, Germany, Japan, the United Kingdom, India, France, Russia, Italy, and Brazil. The study, published in late 2026, brings together two distinct ideas that had rarely been compared side by side. The first is the "exposure" of a country's workforce to artificial intelligence, meaning how many jobs involve tasks that AI can easily perform. The second is the "policy space" available to a government, which is simply the room they have to cut interest rates or increase spending without causing a financial crisis. The researcher built a detailed index for each country, combining data on which jobs use AI with data on how much debt each nation carries and how high their interest rates currently sit. The goal was to see if the countries most threatened by AI were also the ones least able to react to it.

The investigation began by mapping out who actually builds artificial intelligence and who merely uses it. The results revealed a stark imbalance. While the United States and China together produce nearly two-thirds of the total economic output of these ten nations, they are responsible for ninety percent of the private investment in artificial intelligence. In other words, the technology is being created in just two places, while the rest of the world is largely waiting to adopt it. Surprisingly, the countries with the most workers exposed to AI are not necessarily the ones building it. The United Kingdom, the United States, Germany, and France have high levels of exposure, meaning their workers are likely to see their tasks changed by AI soon. Yet, China, which is a massive producer of AI, has a workforce that is relatively less exposed because a large portion of its jobs are in agriculture and basic labor, tasks that are currently less vulnerable to automation. This mismatch means that most of the world's largest economies will experience AI not as creators, but as adopters of a technology developed elsewhere.

The study then tested a specific worry: that the countries most exposed to this disruption would also be the most financially fragile, lacking the money or the interest-rate room to cushion the blow. The data did not support this fear. There was no clear link between how exposed a country's workers were to AI and how much fiscal or monetary room that country had. For instance, Russia and Brazil have high interest rates and low debt, giving them plenty of room to maneuver, while Japan has very low rates and high debt, leaving it with little room. The countries most exposed to AI, like the UK and Germany, have varying levels of financial space, but the exposure itself does not seem to predict whether a country is weak or strong. The research also looked at whether countries that had invested heavily in AI recently were seeing faster economic growth. The answer was no; in the data from 2018 to 2025, higher investment in AI did not correlate with faster growth in how much workers produced. This suggests that the benefits of such a powerful technology take time to appear, or that the investment numbers are not yet large enough to show a clear effect.

The most significant finding emerged from a simulation that asked what would happen if artificial intelligence did successfully boost economic growth. The researcher used a standard economic model to calculate the consequences of a rise in growth for each country's debt and interest rates. The simulation assumed that people do not easily swap their spending between today and tomorrow, a common view in economics. Under this assumption, if AI makes the economy grow faster, the natural interest rate—the rate that keeps the economy stable—must rise even faster. This creates a difficult situation for governments with high levels of debt. Because the interest rate on their debt would rise more than the growth of the economy, the gap between the two widens. To keep their debt from spiraling out of control, these governments would need to run a much larger budget surplus, meaning they would have to collect more in taxes or cut more spending than they currently plan.

This creates a paradox the author calls a "stabilisation gap." On one hand, the rise in interest rates helps central banks. In countries like Japan and the Eurozone, where rates are currently very low and close to the bottom limit, a rise in the natural rate gives them more room to cut rates if a recession hits. It restores a vital tool for managing the economy. On the other hand, this same rise hurts the fiscal side. For countries with high debt, such as Japan, the United States, Italy, and France, the math changes in a way that requires them to tighten their budgets significantly. In a central scenario, Japan would need to increase its primary budget balance by 1.14 percent of its total economic output just to keep its debt stable, while the United States would need to do the same by 0.78 percent. If the economy reacts strongly to the new technology, these numbers could rise even further.

The study concludes that while the countries most exposed to AI are not currently the most financially constrained, the arrival of the technology itself could tighten the constraints. The countries with the highest debt, which are also among the most exposed to AI, face the largest challenge. They will find that the very growth AI brings makes it harder to manage their existing debt, even as it gives their central banks more room to fight recessions. The research does not claim that this future is inevitable, but it highlights a specific arithmetic reality: if artificial intelligence lifts growth, it will likely lift interest rates by more, forcing governments with high debt to make difficult choices. The findings suggest that planning for an AI future requires more than just looking at jobs and productivity; it requires understanding how the technology will reshape the interest rates that govern both monetary policy and public debt.