Petra Tschudin’s message was that AI is not automatically disinflationary: its near-term and medium-term effects could raise prices, while any longer-run productivity benefit may lower costs but does not necessarily produce lower inflation. The SNB therefore treats AI’s net effect as uncertain and will respond to incoming inflation evidence rather than pre-commit to a rate path. 1

• Short to medium term: AI-related investment is being redirected toward some sectors, potentially creating adjustment pressure elsewhere in the economy. Supply bottlenecks—such as shortages of chips—could push component prices higher and add to inflation. 1

• Long term: AI could lift productivity and reduce firms’ costs, creating downward pressure on prices. But Tschudin stressed that this would not mechanically mean deflation; the aggregate effect depends on how investment, demand, supply capacity, and productivity develop. Overall, she said the effect remains unclear. 1

• Related IMF warning: IMF chief economist Silvana Tenreyro likewise cautioned that even a productivity increase from AI need not reduce inflation. The inflation outcome depends on the balance between the demand created by AI investment and the timing and scale of productivity gains. 2

• SNB forecast and the 0% policy rate: The SNB’s conditional inflation forecast placed inflation within its 0%–2% price-stability range through the first quarter of 2029. That is not a promise that the policy rate will remain at its current 0% level for three years; it is a projection conditional on assumptions, including the prevailing policy setting. 1

• Policy implication: Tschudin said the SNB would not simply leave policy unchanged if later data altered the inflation outlook. It remains prepared to adjust monetary policy—including, if necessary, returning to negative rates—to keep inflation within its target range. 1

Petra Tschudin’s message was that AI is not automatically disinflationary: its near-term and medium-term effects could raise prices, while any longer-run productivity benefit may lower costs but does not necessarily produce lower inflation. The SNB therefore treats AI’s net effect as uncertain and will respond to incoming inflation evidence rather than pre-commit to a rate path. 1

• Short to medium term: AI-related investment is being redirected toward some sectors, potentially creating adjustment pressure elsewhere in the economy. Supply bottlenecks—such as shortages of chips—could push component prices higher and add to inflation. 1

• Long term: AI could lift productivity and reduce firms’ costs, creating downward pressure on prices. But Tschudin stressed that this would not mechanically mean deflation; the aggregate effect depends on how investment, demand, supply capacity, and productivity develop. Overall, she said the effect remains unclear. 1

• Related IMF warning: IMF chief economist Silvana Tenreyro likewise cautioned that even a productivity increase from AI need not reduce inflation. The inflation outcome depends on the balance between the demand created by AI investment and the timing and scale of productivity gains. 2

• SNB forecast and the 0% policy rate: The SNB’s conditional inflation forecast placed inflation within its 0%–2% price-stability range through the first quarter of 2029. That is not a promise that the policy rate will remain at its current 0% level for three years; it is a projection conditional on assumptions, including the prevailing policy setting. 1

• Policy implication: Tschudin said the SNB would not simply leave policy unchanged if later data altered the inflation outlook. It remains prepared to adjust monetary policy—including, if necessary, returning to negative rates—to keep inflation within its target range. 1