Artificial intelligence may contribute to higher inflation in the short to medium term, according to Swiss National Bank governing board member Petra Tschudin. In an interview published Friday, Tschudin highlighted risks from redirected capital expenditures and supply bottlenecks—particularly in semiconductor components—that could push prices higher before productivity gains from AI translate into broader economic benefits.
Tschudin noted that while AI-driven productivity improvements have historically not led to structural deflation, the technology’s initial adoption phase often creates imbalances. "Shortages can occur, for example with chips, causing prices to rise," she said. "In the short or medium term, therefore, upward inflationary pressure can also arise."
The SNB’s latest conditional inflation forecast, covering the first quarter of 2029, projects annual price rises within the central bank’s 0% to 2% target range. The projection assumes policy rates remain at their current 0%, though Tschudin emphasized that the forecast does not imply rates will stay unchanged for three years. "The conditional inflation forecast should not be understood to mean that interest rates will remain at their current level for this duration," she said. "We do not publish interest rate forecasts."
Earlier this week, International Monetary Fund chief economist Silvana Tenreyro cautioned in Bank of England research that even productivity gains from AI may not necessarily translate into lower inflation. The divergence underscores the uncertainty surrounding AI’s macroeconomic effects, with potential near-term inflationary pressures contrasting with long-term disinflationary possibilities.
Tschudin added that the SNB would adjust monetary policy if new inflation data warranted it, reiterating the central bank’s data-dependent approach to policy decisions.
