IMF Chief Economist Warns the AI Wealth Boom Could Add to Inflation Pressure
The artificial intelligence boom could add to inflation through more channels than the cost of advanced chips, according to IMF chief economist Pierre-Olivier Gourinchas. In remarks reported by Bloomberg on 26 June, he warned that the AI investment surge is generating very high equity valuations, especially in US markets and in technology-exposed economies such as South Korea, creating a wealth effect that may encourage households to spend more.
The argument broadens the usual macro debate around AI. Most long-term discussion treats AI as disinflationary because it could raise productivity, automate tasks and lower unit costs. Gourinchas is not rejecting that longer-term possibility; he is warning that the transition phase may be more inflationary than markets assume, through two channels. The first is demand-driven: rising AI-related equity values lift household wealth through portfolios and retirement accounts, and wealthier households tend to spend more on services, travel, housing and durable goods, adding demand pressure in an economy where inflation is not fully back to target. The second is supply and input-cost driven: the AI build-out requires enormous spending on data centers, semiconductors, electricity, cooling and grid connections, and that race for capacity can lift demand for chips and power faster than supply adjusts, making AI inflationary in the near term even if it becomes disinflationary later.
Wealth effect versus productivity effect
The key analytical distinction is timing. Productivity gains from AI may take years to diffuse as firms redesign workflows, train staff and change business models, whereas equity wealth effects can operate quickly through rising stock prices and household net worth. That mismatch creates a macro tension: markets may price future productivity immediately, but the real economy can feel the demand impulse before it receives the productivity benefit, so AI can be inflationary in the short run and disinflationary in the long run. The late-1990s technology cycle is an imperfect but useful comparison, when the internet boom lifted investment, valuations and wealth before the full productivity benefits became visible.
The data center channel
The physical infrastructure behind AI is central to the inflation argument. Data centers require land, construction, high-performance chips, electricity, backup power, cooling and network capacity, and as large technology firms race to secure computing power they can create bottlenecks in chip supply, power availability and grid infrastructure. If electricity demand rises sharply in specific regions, local power prices and infrastructure costs can climb; if advanced-chip demand outruns production, semiconductor pricing can stay elevated. AI is a digital technology with a very physical footprint, and that footprint can generate real-economy inflation pressure before AI-based productivity gains become broad enough to offset it.
Why it matters, and the MENA angle
The warning complicates the simple narrative that AI raises productivity, productivity lowers inflation, and central banks can therefore ease. It matters most for the US Federal Reserve, which sits at the center of the AI equity boom, the data-center build-out and the advanced-chip cycle: if wealth effects support consumption while data-center investment supports business spending, the economy may stay more resilient than expected, delaying disinflation and reducing the case for rapid rate cuts. For MENA economies the issue is concrete on the supply side. The region is investing heavily in data centers, cloud capacity and AI infrastructure, which supports diversification and foreign investment but also raises demand for electricity, cooling, land, imported equipment and specialized talent. For energy exporters, AI reinforces a longer-term demand story for power and gas, since data centers need reliable electricity, often from gas-fired generation alongside renewables. The monetary link runs through the dollar: if AI-related demand keeps US inflation sticky and the Fed higher for longer, the Gulf’s dollar pegs import that tightness into bank funding costs, real-estate financing and equity valuations, even where domestic inflation is contained. For sovereign wealth funds, the same boom raises portfolio-concentration questions, as heavy US technology exposure has boosted returns but increases sensitivity to earnings, regulation, power constraints and margins.
Outlook
The key indicators are AI-related equity valuations, data-center capital expenditure, electricity demand, semiconductor pricing, household consumption and productivity data. If AI investment keeps rising while productivity gains stay narrow, inflation risk may run higher than markets expect; if productivity broadens and input bottlenecks ease, the disinflationary case strengthens. The balanced message is that AI may transform productivity over time, but the boom phase can still add to inflation before the efficiency gains arrive, a tension central banks must manage in the short and medium term even as the long-run payoff remains uncertain.
Sources: International Monetary Fund; Bloomberg.
