The Bank for International Settlements (BIS) warns in a report published this Tuesday that the rise of artificial intelligence may distort fundamental indicators for the design of monetary policy, such as natural interest rates and the unemployment rate, thus increasing the danger of a "bad calibration" by central banks.
"The considerable uncertainty surrounding the effects of AI poses several challenges for monetary policy and financial stability," say the authors of the study from the bank of central banks, who emphasize the difficulty of accurately anticipating the impact of this technology.
According to the document, AI is capable of influencing both the supply and demand of the economy simultaneously, both in the cyclical and structural planes, and with very different effects across sectors. This clouds the reading of underlying trends and complicates the task of interpreting the real strength of activity.
Consequently, a phase of strong growth could respond to transient demand impulses and not to a lasting advance in productive capacity, while productivity improvements are distributed unevenly and are difficult to quantify accurately.
"The rise of AI can, therefore, alter key unobservable indicators, such as natural interest rates and unemployment, frequently used in the analysis of monetary policy," warn the economists of the BIS, who insist that this alteration can lead to erroneous diagnoses.
Beyond the size and direction of these effects, the report highlights that they also differ in their "timing": AI could first generate upward pressures on prices and, only later, become a disinflationary force when productivity gains consolidate and supply adjusts.
Thus, although certain short-term inflationary effects, albeit moderate, may already be manifesting, the authors believe that the disinflationary impacts linked to greater productivity or changes in the perception of workers' bargaining power will appear more slowly and gradually.
In this context, if central banks overestimate supply improvements or underestimate the increase in underlying demand, monetary policy can become excessively expansive; while, in the opposite case, more restrictive conditions than necessary could be imposed. Therefore, they argue that a prudent, gradual approach that is heavily data-driven, like the one advocated by the European Central Bank (ECB), helps limit the probability of making diagnostic errors.
The BIS also draws attention to the potential consequences of the AI boom for financial stability. The adoption of this technology can fundamentally alter the business models of certain sectors and provoke negative corrections in their valuations and job prospects, as was already observed in software as a service (SaaS) companies in February 2026.
Furthermore, in some economies, the hefty profits from AI-linked exports could fuel bubbles in domestic asset prices, increasing the vulnerability of the financial system.
More broadly, BIS economists warn that one of the central risks in the current scenario is that expectations regarding the transformative nature of AI may be excessively optimistic. This overconfidence can lead to overinvestment, inefficient capital allocation, and a deterioration in credit quality.
"A correction in asset prices could tighten financial conditions, curb investment, and weaken aggregate demand," conclude the authors of the report, who call for vigilance and caution in the face of the rapid deployment of artificial intelligence in the global economy.