SUMMARY
Artificial intelligence could increase inflation in the short term as massive investment in AI redirects capital and puts pressure on supplies of key components such as semiconductors, Swiss National Bank governing board member Petra Tschudin said. She also warned that while AI could eventually reduce prices by improving productivity, those gains may not automatically create sustained deflation. The SNB is closely monitoring the technology's impact as it assesses future inflation and monetary policy.
ECONOMY & TECHNOLOGY — The rapid expansion of artificial intelligence is often presented as a force that will make businesses more productive and goods cheaper.
But a senior Swiss central banker is warning that the economic impact of AI may be more complicated.
Petra Tschudin, a member of the governing board of the Swiss National Bank, said artificial intelligence could add to inflationary pressure in the short and medium term as investment flows are redirected and shortages emerge in areas such as computer chips.
Her comments challenge the simple assumption that greater productivity from AI will automatically mean lower prices.
AI Could Raise Prices Before Lowering Them
The central issue is timing.
Companies are investing enormous amounts of money in AI infrastructure today, while many of the productivity benefits may take years to fully materialise.
That creates a potential mismatch between demand and supply.
If companies suddenly need more semiconductors, data centres, electricity and specialised equipment, suppliers may struggle to keep up.
When demand rises faster than supply, prices can increase.
Tschudin specifically pointed to possible chip shortages as one example of how AI investment could create inflationary pressure.
Investment Is Being Redirected
AI is attracting huge amounts of capital from investors and companies.
That money does not exist in isolation.
When capital moves heavily toward AI infrastructure, it can reduce the resources available to other sectors or force businesses in those sectors to compete for workers, equipment and financing.
Tschudin said investment flows are being partly redirected, creating adjustments and difficulties for other parts of the economy.
The result could be higher prices in sectors that suddenly face tighter supplies.
Semiconductors Are a Key Example
Modern AI systems require enormous quantities of advanced computing hardware.
That includes high-performance processors, memory chips and other specialised semiconductor components.
When AI companies and data-centre operators compete aggressively for these components, supply can become constrained.
Recent research has also highlighted increases in the prices of some computer components as AI-related demand grows.
This illustrates an important point: technological progress can initially make certain inputs more expensive if demand expands faster than production capacity.
AI Could Eventually Lower Prices
Tschudin does not believe AI is necessarily inflationary forever.
Over the longer term, artificial intelligence could increase productivity, automate tasks and reduce the cost of producing goods and services.
If companies can produce more with fewer resources, those efficiency gains could eventually translate into lower prices for consumers.
That is the strongest argument for the long-term deflationary potential of AI.
But Tschudin cautioned against assuming that productivity gains automatically produce sustained deflation.
Why Productivity Does Not Automatically Mean Deflation
Inflation measures how prices change over time.
For productivity improvements to create persistent deflation, prices would need to continue falling repeatedly rather than simply becoming cheaper once.
Tschudin questioned whether that is realistic.
Productivity improvements have existed throughout economic history, yet they have not automatically pushed economies into long-term deflation.
Other forces—including wages, demand, investment and government policy—continue to influence prices.
The AI Boom Is Becoming a Macroeconomic Issue
AI was once primarily viewed as a technology story.
That is changing.
The scale of investment in data centres, chips, electricity infrastructure and AI software means the technology is increasingly affecting the broader economy.
Central banks now have to consider whether the AI boom is creating additional demand, increasing productivity or producing shortages.
The answer could affect interest-rate decisions in countries around the world.
The Swiss National Bank Is Watching Closely
The Swiss National Bank is closely monitoring AI's impact on prices.
The central bank's current conditional inflation forecast places Swiss inflation within its price-stability range of 0% to 2% through the first quarter of 2029, assuming its policy rate remains unchanged.
But Tschudin stressed that the forecast should not be interpreted as a promise that interest rates will remain unchanged.
If new inflation data changes the central bank's assessment, monetary policy can change.
Switzerland's Interest Rate Is Already Very Low
The SNB's policy rate currently stands at 0%.
That gives the central bank limited room to respond through conventional rate cuts if inflation were to weaken significantly.
At the same time, if inflationary pressure were to rise unexpectedly, the SNB could face a different challenge: deciding whether tighter monetary conditions are necessary even while economic growth remains uncertain.
The AI boom therefore creates another variable for policymakers to monitor.
AI Spending Could Keep Driving Demand
The world's largest technology companies continue to invest heavily in AI infrastructure.
Those investments require physical resources.
Data centres require land, construction materials, electricity, cooling systems and networking equipment.
AI models require increasingly powerful computing hardware.
All of those inputs can experience supply constraints if investment expands too quickly.
This is one reason the AI boom can produce inflationary pressure even if the technology ultimately makes other parts of the economy more efficient.
Energy Could Become Another Bottleneck
Semiconductors are not the only potential constraint.
Large AI data centres consume enormous amounts of electricity.
As more facilities are built, demand for power generation and grid capacity can rise sharply.
If electricity supply does not expand at the same pace, energy prices can come under pressure.
This creates another pathway through which AI investment could influence inflation.
The Labour Market Could Also Change
AI could have complicated effects on wages and employment.
Some tasks may become cheaper to perform because AI can automate parts of knowledge work.
At the same time, demand could increase for workers who design, operate and maintain AI systems.
If specialised workers become scarce, their wages could rise.
The ultimate effect on inflation will depend on which forces become stronger.
The IMF Is Raising a Similar Question
Tschudin's warning comes shortly after research involving new International Monetary Fund chief economist Silvana Tenreyro also questioned whether AI-driven productivity gains will necessarily reduce inflation.
The research argued that investment and consumer spending can increase before the productivity benefits of AI are fully realised.
That temporary imbalance can increase demand and put pressure on prices.
The researchers also noted that productivity gains do not have identical effects across different sectors of an economy.
The Long-Term Question Remains Open
The biggest uncertainty is what happens after the current investment boom.
If AI eventually makes production dramatically more efficient, the technology could reduce costs across many industries.
But if demand for AI infrastructure continues expanding faster than supply, inflationary pressures could persist for longer.
Central banks therefore cannot assume that AI will automatically solve inflation.
Why This Matters Beyond Switzerland
The Swiss economy is not the only one facing this question.
The United States, Europe and Asia are all investing heavily in artificial intelligence.
Many of the same constraints—chips, electricity, data centres, skilled workers and infrastructure—are global.
If AI investment produces widespread shortages, the inflationary effects could spread across borders.
Conversely, if productivity gains become large enough, the technology could eventually lower production costs worldwide.
AI Could Create Both Inflation and Deflation
The most realistic conclusion may be that AI can push prices in both directions at different times.
In the short term, investment can increase demand for scarce resources.
In the longer term, automation and productivity improvements can increase supply and lower production costs.
The timing between those two effects will determine whether AI becomes inflationary, deflationary or broadly neutral.
Our Perspective
The biggest mistake would be to assume that technological progress automatically makes everything cheaper.
Technology can reduce costs, but the transition itself can be expensive.
AI is a perfect example.
Companies are spending enormous sums today on hardware, electricity, data centres and talent in the expectation of future productivity gains.
That spending can create shortages before the promised efficiency arrives.
AI may eventually make the economy more productive, but the journey toward that productivity could itself create inflationary pressure. Central banks therefore have to watch not only what AI can produce, but also what the AI boom is consuming.
Conclusion
Artificial intelligence could push inflation higher in the short term as investment flows shift toward AI and demand for scarce components such as semiconductors increases, according to Swiss National Bank governing board member Petra Tschudin.
At the same time, AI has the potential to lower prices over the longer term by improving productivity and making goods and services cheaper.
The challenge for policymakers is determining which effect will dominate and how quickly.
The SNB is therefore monitoring the technology as part of its assessment of future inflation and monetary policy.
The AI revolution could ultimately be a powerful force for lower costs—but in the short term, the enormous investment required to build it may have the opposite effect.
Daily Touch Insights Editorial Team
View Journalist Profile
