ROI-On AI and Iran, Trump risks bombing himself in the foot: Joachim Klement

US President Donald ​Trump’s administration has made the AI boom a signature economic bet. A strike on Iran after the US midterm elections could ​potentially blow that up. The White House has doubled down on its support for the artificial ‌intelligence ​revolution, claiming that the country that wins the AI arms race will win the future.

Yet, at the same time, the US government has indicated that it may attack Iran again after Tuesday's midterms if no new ceasefire deal is reached. The White House's signalling has been mixed in recent days, but, if pursued, this hawkish foreign policy threatens to undermine Trump’s AI vision. That’s because a renewed Middle East ‌conflict could intensify the current oil shock, keep inflation elevated and force the Federal Reserve and other central banks to tighten policy aggressively, pushing up funding costs for the data-centre buildout fuelling the AI boom.

BRENT BOMB Energy markets are at the heart of all this. Fears of renewed US-Iran fighting, a renewed effective closure of the Strait of Hormuz or further attacks on Saudi energy infrastructure already drove crude oil prices back to their March highs late last week, above $115 a barrel.

While Brent crude prices fell more than 8% on Monday amid optimism about a potential diplomatic resolution and slid further ‌on Tuesday following news that the US would cut military aid to Israel, the risk to prices is still to the upside given how much space remains between Washington’s and Tehran’s positions. Rising energy prices – or simply persistently elevated prices – would complicate central-bank policy decisions significantly. ‌If oil were to fall back below $100, favourable year-on-year comparisons in the second quarter of 2027 would help pull energy inflation sharply lower. That's because prices would be measured against the high levels recorded in the second quarter of 2026 after the Iran war began. That would not only lower headline inflation but also likely ease pressure on core prices as transport, heating and cooling costs would be expected to decline.

But if oil remains near recent highs or rises further, that relief will not materialise. Persistently high energy inflation could spill into core prices, delaying expected rate cuts or forcing policymakers to tighten policy more aggressively than markets currently expect. While higher rates may not cool the broader economy ⁠until later in 2027, ​they could strike an immediate blow to the AI boom.

A HIGHER HURDLE FOR ⁠AI The massive AI data centre buildout requires huge upfront spending on chips, power, land and construction. Over the last two months, investors have become increasingly concerned about the scale of that capex, as evidenced by the negative stock market reaction to recent announcements of big spending plans from the likes of Alphabet and Amazon.

The big five hyperscalers — ⁠Alphabet, Amazon, Apple , Meta and Microsoft — have already exhausted their cash piles and turned free-cash-flow negative as a group, though Microsoft continues to generate free cash flow. These firms are therefore relying increasingly on debt to finance further capex. Yet their spreads over 5-year US Treasuries have widened significantly over the past year, raising ​the return hurdle that AI infrastructure projects must now meet.

Neoclouds, which specialise in supplying GPU computing power to AI firms, are more exposed still. They lack the hyperscalers' large and established cash flows and already face exceptionally high borrowing costs. CoreWeave, for example, ⁠has effectively lost access to viable debt financing. With borrowing costs above 15%, new data-centre projects are difficult to justify economically. That problem will only get worse if the Fed keeps hiking interest rates in response to an energy-driven inflation threat, as this will lift 5-year Treasury yields as well.

Hyperscalers and neoclouds would then face a twofold setback: wider credit spreads ⁠driven ​by company-specific balance sheet concerns and higher benchmark rates caused by persistent inflation. THE NEGATIVE FEEDBACK LOOP

This could ultimately create a negative feedback loop: investor concerns about the AI capex boom could push up borrowing costs, fuelling further doubts about profitability, slowing investment, and thus pushing down hyperscaler and AI-sector share prices. The entire AI bull case seems to rest on rapid revenue growth that can justify past and future capex. If revenues slow sufficiently, some projects may simply become unprofitable, and the entire narrative could fall apart.

Obviously, none of this is cast in stone. This scenario depends ⁠on an escalation in the Middle East conflict, sustained higher oil prices and a forceful policy response from central banks. If a diplomatic solution is reached, energy supplies remain intact or policymakers look through a temporary inflation spike, funding costs for hyperscalers and neoclouds ⁠may rise only slightly, or not at all. But for an administration eager to ⁠keep the AI boom going, it is ironic that it is contemplating taking an action that could slow or even break that boom. Instead of shooting itself in the foot, Washington risks bombing itself where it hurts.

(The views expressed here are those of Joachim Klement, an investment strategist for Panmure Liberum.) Enjoying this column? Check out Reuters Open Interest (ROI), your essential new source for global financial commentary. Follow ROI on ‌LinkedIn, and X.

And listen to the Morning Bid ‌daily podcast on Apple, Spotify, or the Reuters app. Subscribe to hear Reuters journalists discuss the biggest news in markets and finance seven days a ​week. (Writing by Joachim Klement Editing by Marguerita Choy and Anna Szymanski)