New York open: AI’s inflation engine is running hot beneath the market’s calm surface

AI’s inflation engine is running hot

On the surface, global markets appeared remarkably calm as Tuesday’s New York session began, but beneath that placid exterior, two powerful currents continued to collide. One was the familiar geopolitical drama surrounding Iran and the Gulf. The other was the increasingly dominant AI investment boom, quietly reshaping everything from corporate financing and factory activity to inflation expectations and monetary policy. The irony is that while traders remain glued to every fluctuation in crude oil prices, the bigger inflation story may be emerging from server racks, semiconductor fabs, and data centres rather than from oil tankers navigating the Strait of Hormuz.

The latest chapter in the AI saga arrived with the force of a fresh gust of speculation. Anthropic confidentially filed for an IPO, appearing to jump ahead of OpenAI and positioning itself directly in the slipstream of the highly anticipated SpaceX offering expected later this month. Taken together, these deals are not merely corporate fundraising exercises. They are potential tectonic events for equity markets. Anthropic’s latest funding round values the company at nearly $1 trillion, while SpaceX is preparing an offering that would place its valuation near $1.75 trillion. Those numbers are no longer venture capital curiosities. They are large enough to alter benchmark composition, shift index weightings, and further concentrate market leadership into an increasingly narrow group of AI-related giants.

At the same time, Alphabet added another wrinkle by announcing roughly $80 billion in new equity financing, including a sizeable private placement with Berkshire Hathaway. That move may ultimately prove more significant than any individual IPO. Investors have become accustomed to hyperscalers issuing mountains of debt to fund the AI arms race. Equity issuance is a different beast altogether. Debt markets absorb borrowing. Equity markets absorb ownership dilution. The question now confronting investors is whether demand for AI exposure remains deep enough to digest an ever-expanding pipeline of stock offerings at valuations that already assume extraordinary future success. Markets often resemble a theatre where everyone rushes through the same entrance. The problem is rarely getting in. It is discovering how narrow the exit can become.

Yet unlike many past speculative episodes, this boom is not operating entirely on dreams and narratives. The physical demand is increasingly visible. Europe’s STMicroelectronics surged to its highest level since 2000 after doubling its data center revenue outlook for the year. That is not the language of speculation. That is the language of real orders, real customers, and real infrastructure spending. The AI gold rush is creating an entire ecosystem of picks, shovels, railroads, and supply depots. Every new model requires chips. Every chip requires fabrication. Every fabrication plant requires power, cooling, software, networking equipment, and substantial capital. The investment machine has become self-reinforcing.

That is precisely why I suspect investors may be watching the wrong inflation dashboard. Every central banker on the planet is monitoring oil prices for signs of another cost-of-living shock. Brent’s recent volatility reflects that concern. Prices surrendered part of Monday’s surge after President Donald Trump suggested negotiations with Iran could continue, yet the broader picture remains striking. Even after the recent pullback, year-end crude contracts remain more than 30% above prewar levels. Markets are attempting to price not today’s headlines but tomorrow’s probability tree. The challenge is that every diplomatic breakthrough seems to arrive attached to another geopolitical tripwire.

Still, oil may ultimately prove to be the shorter-lived inflation impulse. The AI buildout appears far more persistent. Capital spending directly tied to artificial intelligence is expected to exceed $800 billion this year alone, with cumulative investment likely to run into the trillions over the coming years. That spending is already influencing earnings, economic growth, industrial production, and increasingly, prices. While public debate continues to focus on whether AI will eventually replace workers and create long-term deflationary forces, much less attention has been paid to what comes first. Before technology lowers costs, it often raises them.

Evidence of that dynamic is beginning to emerge in unexpected places. One of the more intriguing inflation signals has appeared inside software and computer accessory pricing. Recent inflation data showed an extraordinary surge in this category, accounting for an outsized share of gains in core goods inflation. A significant driver appears to have been memory-related products, particularly as global data center construction fueled a shortage of memory chips centred in South Korea. In simple terms, the digital economy is consuming physical resources at a pace that supply chains struggle to keep up with. Before AI becomes a productivity miracle, it first becomes a demand shock.

Some economists argue these figures may overstate the true inflationary impact. Statistical agencies often struggle to measure rapidly evolving technologies. Quality improvements, changing product specifications, and category mismatches can create distortions. That may be true. But traders should never confuse imperfect measurement with an absence of pressure. Even if the precise numbers are debatable, the underlying forces are unmistakable. Memory prices have risen. Semiconductor demand has exploded. Infrastructure spending is accelerating. These are observable realities rather than statistical abstractions.

That distinction matters because the market’s inflation assumptions increasingly rest on the belief that AI will eventually become disinflationary. Perhaps it will. But markets often underestimate the sequence of events. Massive investment booms rarely arrive without bottlenecks. They rarely arrive without shortages. They rarely arrive without pricing power emerging somewhere in the supply chain. In many ways, the current AI cycle resembles an industrial revolution unfolding at internet speed. Demand is arriving immediately while productive capacity takes years to catch up.

The result is that AI may be exerting upward pressure on interest rates rather than downward pressure. Not because it is destroying productivity, but because it is generating extraordinary demand throughout the economy. The market increasingly resembles a giant construction site where every contractor needs the same materials at the same time. Chips, memory, software, electricity, engineering talent, and financing are all being pulled into the same vortex.

This is where the inflation story becomes more nuanced. The danger is not simply higher prices. It is the possibility of a sequence familiar to students of economic history. First comes overheating. Then comes policy tightening. Then comes cooling. In other words, heat followed by cold. If AI-related investment continues accelerating while energy markets remain constrained, central banks may find themselves confronting a double inflation shock from two very different directions. One arrives through oil pipelines. The other arrives via fibre-optic cables.

For casual investors, this matters because even to this veteran trader, cross-asset correlations are confusing as hell. Equities continue to celebrate growth. Bond yields continue to signal inflation concerns. Oil continues to signal supply risk. The dollar remains firm. These are not messages that normally coexist comfortably. The market orchestra is still playing, but the instruments are no longer following the same sheet music.

For now, the AI boom remains the dominant melody. Yet history teaches that every great investment mania eventually confronts the same question. Not whether the technology is real, but whether the market has already paid for too much of the future. As billions become trillions and speculative excitement evolves into unprecedented equity issuance, investors would be wise to remember that bubbles rarely burst because the story was false. They burst because the price of believing it became too high.

That is why the real inflation battle may not be fought in the Gulf at all. It may be fought inside data centers humming away across America, Europe, and Asia. Oil remains the headline risk. AI may be the structural one. And unlike a geopolitical ceasefire, there is no obvious peace treaty coming for an investment boom that has only just begun.