Economists Warn of Major Inflation Next Year, Here Are the Culprits

The United States and Iran have agreed to de-escalate the war in the Middle East. Additionally, US President Donald Trump has begun to soften his stance in the trade war against China. World oil prices have consequently started to show a downward trend. However, inflation fears still loom. Economists point to a new catalyst: the massive development of artificial intelligence in the US. The phenomenon of technology giants doing whatever it takes to win the AI race is expected to impact price increases, ranging from consumer electronic devices such as mobile phones and laptops, to electricity tariffs. The money flowing into the AI race is unprecedented. Analysts estimate capital expenditure at five companies known as hyperscalers—Alphabet, Amazon, Meta Platforms, Microsoft, and Oracle—could reach US$741 billion this year, according to FactSet. That figure is up nearly 75% from last year. Although much of the conversation focuses on what AI can do, the build-out itself is intensely physical, said Columbia University economist. The data centres used for AI require advanced computing equipment, cooling systems to prevent the equipment from overheating, power and fibre-optic cables, and backup generators to prevent power disruptions. Based on announced and planned developments, spending on AI build-out through 2032 could reach around US$8 trillion, five times the market value of the entire New York City property market. The high demand for AI will raise the prices of a string of products and services needed to build AI. Many of these products and services are also used for other purposes, not just AI. As a result, these price increases impact the broader economy. Take memory and storage chips used in a wide range of consumer electronics, from video game consoles, mobile phones, and laptops, to cars. Nintendo, Microsoft, and Sony have all raised the prices of their devices. Higher prices will also apply to Apple products, according to CEO Tim Cook, who told The Wall Street Journal that the cost surge is unlike anything he has ever seen in any field in over 40 years. If AI truly is as revolutionary as many economists predict, it could eventually cool inflation. That is the lesson from past technological revolutions, which boosted worker productivity, making it easier for businesses to meet demand without raising prices. Kevin Warsh, who now serves as the Federal Reserve chair, has previously made this point. ‘AI will be a significant disinflationary force, boosting productivity and strengthening American competitiveness,’ he wrote in the Journal in November 2025. Warsh’s views on the inflationary impact of the AI boom will be a major first test of his leadership at the Fed. AI infrastructure can be built much faster than transformative technologies of the past, such as railways in the 19th century, electrification in the early 20th century, or telecommunications during the dot-com boom. But it takes a long time to bear fruit. Even on an accelerated timeline, economists at UBS estimate it will take at least a few years before AI starts helping to lower inflation. In the short term, AI-driven demand is causing higher prices. In a National Association for Business Economics survey of economists released earlier this week, 81% of respondents said the AI build-out will add to inflation over the next year. ‘In the first phase of any major technological revolution, you tend to experience pressure on scarce resources, and that tends to put upward pressure on prices,’ said EY-Parthenon chief economist Gregory Daco, who is NABE president. This is already starting to show up in inflation data. According to the US Department of Labour, consumer prices for computer software and accessories rose about 15% year-over-year in May. Further price increases are likely: The Labour Department’s measurement for wholesale electronic components and accessories rose 27% year-over-year last month. According to strategists at Evercore ISI, the impact of AI development on prices could be very different from the tariffs imposed last year or this year’s fuel price surge. Both tariffs and oil are one-time shocks, which temporarily affect price increases. AI is a demand shock that could persist for years. Indeed, most of that demand shock has yet to arrive. Fed Governor Lisa Cook noted in a speech last month that only a small fraction of the announced spending on data centres has been deployed. The money OpenAI and Anthropic are expected to raise in their upcoming initial public offerings could add more fuel to the AI build-out. This dynamic is reflected in the rally of chip company stocks, which have moved sharply higher on investor expectations of a sharp increase in demand. Even with this week’s sharp sell-off, the PHLX Semiconductor Index is up about 150% over the past year. Of course, it is not just chips that go into data centres. Many other things go into building and operating data centres, which are also widely used across the economy. This can also raise costs for various businesses, which may then try to pass them on to customers.