The artificial intelligence is no longer knocking at the door of the economy: it is already inside, and for the central banks means dealing with a transformation which can change demand, inflation, jobs and markets. For Fabio Panetta, Governor of the Bank of Italy, is one of the main challenges for monetary policy, along with growing geopolitical fragmentation. "AI will change the economy and markets, so central banks cannot stand by and watch."
Panetta's speech at the international conference of the National Bank of Ukraine in Kyiv starts from the need to adapt tools and analysis to an economy in which artificial intelligence is spreading faster than previous technological innovations. "Its adoption has surpassed that of personal computers and the internet at comparable rates."
AI, demand, and inflation: what changes for central banks, according to Panetta
The impact of AI on monetary policy, however, will not necessarily follow only one direction. On the one hand, the investments The investments needed to develop artificial intelligence are already supporting demand. "AI-related investments are already providing a strong boost to demand." If supply fails to keep pace, this imbalance could translate into increased price pressures: "These pressures could push up relative prices and inflation."
On the other hand, the main effect of AI could come from productivity. “The dominant effect will likely be in terms of productivity. The extent of the benefits remains highly uncertain, but their potential is considerable.” For Italy, analyses by the Bank of Italy indicate that widespread adoption of artificial intelligence could increase labor productivity growth. over one percentage point per year.
The problem for central banks is to understand when and with what intensity These effects will manifest themselves. "The timing, magnitude, and transmission of these effects are highly uncertain." Therefore, according to Panetta, institutions will need to "refine their analyses" to better capture the consequences of AI on demand, consumption, "inflation forecasts," and the banking system.
AI, work, and consumption: two possible paths
Making the picture even more complex is the job marketAutomation can change employment and wages and, consequently, the way families spend and save. If uncertainty about the future occupational, in fact, prudence may prevail. “Greater uncertainty regarding jobs and wages could induce families to save more and consume less. A drop in consumption could therefore partially offset the boom in investment, allowing the disinflationary effects of AI to emerge earlier.”
But there's also another scenario. If artificial intelligence creates new activities e supports income from workExpectations may change. "If artificial intelligence creates new tasks and increases expected labor income, families may feel wealthier and more confident about the future." In this case, "consumption could strengthen" and "inflationary pressures could persist longer."
The transition, in short, could be "anything but smooth," because the effects of AI on productivity, employment, demand, and prices could manifest themselves simultaneously but in different directions and at different times.
United States and Europe: Speed of adoption can change competitiveness
Another element concerns the different capacity of the economies of adopt and exploit artificial intelligenceAccording to the head of Via Nazionale, "differences in adoption rates could have significant consequences for growth."
The issue isn't just about who develops the most advanced AI models, but also the ability of businesses and economic systems to utilize these technologies. "Even if Europe doesn't become a leader in AI development, it has the potential to adopt and exploit these technologies as rapidly as other advanced economies."
Le consequences could be relevant in terms of productivity and competitiveness international: "Integrating AI more rapidly could gain a lasting productivity advantage, expand their share of global markets, and attract greater investment and capital. Others, however, risk a decline in their relative competitiveness."
The spread of AI could therefore contribute to change the economic balance between countries, creating new differences not only in productivity growth, but also in the ability to attract capital and conquer market share.
AI, trade and exchange rates
The consequences of technological transformation could extend beyond monetary policy and the labor market. Differences in productivity and demand would have repercussions on international economic relations. "Disparities in productivity and demand would also affect trade flows, capital movements, and relative prices. Exchange rates would be part of this adjustment process and could experience greater volatility during the transition phase," Panetta explains.
It is an aspect that broadens the perimeter of the challenge for central banks. AI could in fact affect monetary policy “not only through productivity and the labor market, but also through the terms of trade and exchange rates”.
For this reason, according to Panetta, the actual developments in demand, economic activity and inflation should have a greater weight in guiding monetary policy decisions compared to the estimates of the neutral rate, that is, the level of rates considered compatible with an economy in equilibrium.
In a context characterized by rapid transformations, central banks “must move” pragmatically, based on data and taking into account an environment in which the effects of innovation are still difficult to “evaluate”.
Economic models and financial system to be updated
The speed of technological transformation also poses a problem More directly: the tools used by institutions were built in a very different context than the current one. "Our economic models, supervisory tools, and payment infrastructures were primarily designed for a different technological context. We must understand the changes underway, distinguish structural transformations from temporary effects, and adapt our tools accordingly."
The update It's not just about macroeconomic models, but also about central banks' ability to monitor the risks posed by the use of artificial intelligence in the financial system.
A possible critical point is the concentration tecnologica. “If many banks rely on similar models or a small number of technology providers, errors or disruptions could ripple throughout the financial system.” The adoption of AI can therefore increase efficiency, but it also introduces new forms of interdependence that authorities will need to be able to identify.
Markets and the risk of corrections
The transformation of the economy is already reflected in the financial assessments. Current asset prices reflect optimistic expectations about the future profitability of AI. These valuations help technology companies raise capital and support further investment, but they also make markets vulnerable to sharp corrections when expectations are not met, as recent volatility in technology stocks has demonstrated.
The point, then, is not only how much AI will actually be able to increase productivity, but also how much expectations of investors are already embedded in asset prices. If economic results were to fall short of expectations, rapid market corrections could emerge.
Central banks' credibility under pressure
The technological issue is finally part of a broader challenge for central banks, which Panetta also links to fragmentation geopolitics. “Geopolitical fragmentation and technological change pose very different challenges for central banks, but they raise the same fundamental question”: “How do we maintain credibility when the environment around us is profoundly changing?”
The answer, according to Panetta, isn't built when a crisis strikes. "Credibility isn't earned when a crisis hits"; "it's built over time through solid institutions, rigorous decisions, and the willingness to undertake difficult reforms. It's then put to the test when circumstances become extreme."
AI thus represents “a different kind of proof” for monetary institutions: does not require abandoning the principles underlying monetary policy, But say adapt tools, skills and analytical capabilities to a rapidly changing economy. "Adapting, however, does not mean abandoning our foundations. Price stability, financial stability, and trust in money remain our anchors."
