Artificial intelligence is moving from being primarily a technology story to becoming a factor central banks may need to consider when assessing inflation.
European Central Bank Governing Council member Fabio Panetta said on September 21 that policymakers need to understand who ultimately benefits from AI-driven economic gains because their distribution could influence aggregate demand and, in turn, inflation.
The issue is becoming increasingly relevant as AI investment expands and the technology begins affecting productivity, labour markets, financial markets and payment systems. But the direction of its impact on prices is far from straightforward: AI could create inflationary pressure through stronger spending and investment while also eventually lowering costs by improving productivity and expanding economic capacity.
Why the Distribution of AI Gains Matters
Panetta's argument highlights an important distinction in the AI debate: economic growth alone does not determine what happens to inflation.
If AI-driven productivity gains translate into higher incomes and stronger spending, aggregate demand could rise. Businesses may also increase investment as they deploy AI infrastructure, software and computing capacity.
But the inflationary impact could be weaker if a large share of the gains is concentrated among groups with a lower tendency to spend additional income.
ECB Executive Board member Philip Lane made a related point in a March analysis, noting that increased income and wealth inequality could limit the expansion of demand and therefore reduce some of the inflationary pressure associated with AI productivity gains.
This means central bankers may increasingly need to examine where AI-generated income and wealth are going, rather than looking only at headline productivity figures.
AI Could Push Inflation Higher — At Least Initially
There are several channels through which rapid AI adoption could add to price pressure.
Companies developing and deploying advanced AI systems require substantial computing infrastructure. That means investment in data centres, specialised chips, networking equipment and electricity generation or supply.
Lane has argued that building this computational infrastructure requires significant upfront capital expenditure. Growing demand for computing power could also increase energy consumption, potentially putting upward pressure on energy prices until additional supply becomes available.
Another ECB analysis published in March said AI could be more inflationary than disinflationary in the short term because of heavy investment in energy-intensive data centres and possible bottlenecks involving specialised chips and skilled workers.
So even if AI ultimately makes businesses more efficient, getting to that point could itself create new sources of demand.
But Productivity Could Eventually Lower Price Pressures
The opposite mechanism is equally important.
If AI allows workers and companies to produce more with the same amount of labour and capital, the economy's productive capacity could increase. Greater supply, all else equal, can reduce pressure on prices.
There is already evidence of substantial productivity improvements in specific applications, although translating those results into economy-wide productivity growth remains uncertain.
Lane cited research in which access to ChatGPT reduced the time required for certain professional writing tasks by 40% while improving output quality by 18%. Another study involving more than 5,000 customer-support agents found a 15% average increase in issues resolved per hour after deployment of a generative-AI assistant.
The ECB cautions, however, that such micro-level improvements do not automatically mean similarly large productivity gains across the entire economy.
Different industries have different opportunities to automate or augment work, while implementation costs, skills shortages and organisational changes could slow adoption.
AI Is Already Becoming Part of the Euro-Area Growth Story
The debate is no longer entirely theoretical.
In its September 2026 macroeconomic projections, the ECB said the euro-area economy had proved more resilient than expected to the shock caused by the Middle East conflict. Among the factors supporting domestic demand, it specifically identified AI-related investment, alongside a robust labour market and public investment.
At the same time, the ECB projects annual inflation at 3.0% in 2026, with inflation expected to peak late in the year as the Middle East conflict pushes energy prices higher. Inflation is projected to decline to 2.1% in 2028.
That makes distinguishing AI-related inflationary effects from energy and other shocks especially important for policymakers.
ECB Says Economy-Wide AI Inflation Effects Remain Unclear
Despite the rapid expansion of AI, there is an important caveat.
ECB research published this month said it does not yet find evidence of a general euro-area demand boost from private investment that can be attributed to AI, although effects may already be significant in individual countries or sectors.
That distinction matters.
The ECB is therefore not saying that AI is currently a major cause of euro-area inflation. Rather, policymakers are increasingly examining AI as a structural force capable of influencing inflation as investment, productivity and adoption grow.
Earlier ECB analysis similarly concluded that AI's aggregate effects on euro-area productivity, employment and inflation remained limited and uncertain, with the eventual outcome depending heavily on the speed and breadth of adoption and the investment required to implement the technology.
What AI Could Mean for Interest Rates
AI could eventually influence another crucial variable for central banks: the economy's natural rate of interest.
If businesses expect major productivity improvements, they may invest more heavily. Consumers anticipating higher future incomes may also save less and spend more.
Lane's analysis suggests that sustained optimism about AI-driven productivity and income gains could increase investment and reduce savings, putting upward pressure on the natural interest rate.
That could have important implications for monetary policy because it would potentially change the interest-rate level consistent with stable economic activity and inflation.
The ECB has already acknowledged the broader challenge. Its updated monetary-policy strategy identifies the increasing use of AI, alongside geopolitical fragmentation, demographic change and environmental pressures, as structural shifts that could make the inflation environment more uncertain and potentially more volatile. The ECB continues to target 2% inflation over the medium term.
Why This Matters
The emerging AI-inflation relationship is more complicated than the idea that automation simply lowers prices.
AI can simultaneously increase productivity, encourage corporate investment, boost household income, raise electricity consumption, increase demand for scarce computing resources and alter the distribution of wealth.
Some of those forces are inflationary; others are disinflationary.
The eventual outcome will depend on how quickly AI spreads, how much new investment it generates, whether productivity improvements become economy-wide, and who ultimately receives—and spends—the resulting economic gains. That uncertainty explains why the ECB is treating AI as an increasingly important variable to monitor rather than assuming it will automatically push inflation in one direction.
