AI developers have been exceptionally fortunate—their technological breakthroughs have coincided with a financial system ideally positioned to fund their ambitious expansion plans. During this period, interest rates were relatively low and declining, energy prices were cheap, and global markets, including tech giants, were flush with cash. Even before the outbreak of the Iran war, some of these favorable conditions had begun to shift, but last week, they underwent a dramatic transformation.

As the conflict enters its second week, economic headlines are dominated by oil and gasoline prices. These commodities are not directly linked to AI infrastructure development—oil is rarely used for power generation, and gasoline costs are negligible for data center developers.

However, secondary effects are significant. Rising oil prices could drive inflation higher, which typically leads to increased interest rates. Both factors are highly unfavorable for AI construction projects burdened with debt, as companies would face higher building costs and interest expenses.

Persistent inflation has also completely dashed market expectations for further Federal Reserve rate cuts this year. Previously, investors anticipated that the Fed would lower rates later in the year as inflation eased and the labor market weakened further. However, the risk of rebounding inflation is now increasing. The U.S. job market has already shown signs of deterioration—nonfarm payrolls fell by nearly 100,000 in February.

If the war continues, it is likely to hamper economic growth. The biggest concern on Wall Street currently is stagflation—a scenario characterized by weak economic activity coupled with high inflation, which would constrain the Fed's ability to cut rates.

None of these developments bode well for large-scale infrastructure construction. Should the conflict prolong, the already substantial U.S. government deficit is expected to widen further, potentially driving interest rates even higher.

Even before the war, AI infrastructure development faced political and financial challenges, and these are likely to intensify. Although data centers do not consume large amounts of oil or gasoline, ordinary consumers do. Meanwhile, a significant portion of data center electricity comes from natural gas power generation. Consumers have already blamed rising electricity prices on data centers, so it is not difficult to foresee that increasing fuel costs will trigger more opposition to AI construction. Local resistance across many parts of the U.S. has already made building data centers increasingly difficult.

The timing of this fuel price surge is particularly unfortunate. A severe winter across much of the northern United States has driven heating costs sharply higher, further squeezing household budgets. According to the National Energy Assistance Directors Association, heating expenses were already projected to rise by 11% this season, and the cold weather has pushed costs even higher.

An ironic scene unfolded last week: just days after President Trump initiated military action against Iran, he summoned the heads of seven major AI developers to Washington, promising to keep consumer electricity prices low. The event was largely political theater. Companies such as C3.ai, Inc., Amazon.com, and Microsoft have already taken significant steps to reduce electricity costs.

More ironically, these companies have not adhered to Trump’s opposition to using renewable energy for AI operations. They have opted for solar, wind, and battery storage because these are the lowest-cost and most accessible forms of energy. (Do not be misled by this week’s dip in oil prices; industry veterans warn that if oil shipments through the Persian Gulf remain disrupted, serious challenges lie ahead.)

Funding channels for the AI sector remain open for now. Shares of major AI developers have held steady, and planned initial public offerings for SpaceX, OpenAI, and Anthropic are still proceeding. The main cracks are appearing in private credit markets, where smaller investors are pulling out of large lenders such as Blue Owl and Blackstone. This is less related to the war: investors are concerned about AI construction costs and lenders’ exposure to software companies, which they view as vulnerable to AI disruption. Nevertheless, rising interest rates will further squeeze these funding sources.

Tech giants are more directly impacted by the war itself due to their operations in the Middle East, rather than general market turbulence. Last week, three of Amazon.com’s data centers in the region were attacked by drones, and the conflict could also jeopardize several large data center projects planned there. Despite this, Amazon.com successfully issued approximately $50 billion in bonds in the U.S. and Europe this week, with the offering being oversubscribed.

AI infrastructure development has previously benefited from easy financing conditions and the strong financial positions of tech giants. For now, both pillars remain intact. However, wars often have a way of overturning established dynamics.