PUBLISHED: March 7, 2026 | LAST UPDATED: March 7, 2026

We went from "quick operation" to "wars can be fought forever" in less than one week and almost nobody is connecting that timeline to a trillion-dollar industry quietly sitting on a foundation this conflict is shaking apart.

On February 28, 2026, the US and Israel launched Operation Epic Fury against Iran. In the first 48 hours, CENTCOM struck more than 1,000 targets, double the scale of Shock and Awe in Iraq in 2003, according to Bloomberg. Supreme Leader Khamenei was killed. Iranian air defenses were systematically dismantled.

And yet the war is not wrapping up. Trump said two to three days, then four to five weeks. At the first Pentagon press briefing since the strikes, Hegseth stood alongside General Dan Caine and declared: "This is not Iraq. This is not endless. This operation is a clear, devastating, decisive mission, destroy the missile threat, destroy the navy, no nukes." Pressed on the actual timeline, he added it "could move up, could move back, four weeks, two weeks, six weeks." JD Vance insisted Trump would not "allow" a long war because he "has clearly defined what he wants to accomplish."

Netanyahu stated the goal was to "remove the existential threat posed by the terrorist regime in Iran" and create conditions for "the brave Iranian people to take their destiny into their own hands." Days later, an internal CENTCOM assessment leaked to Middle East Eye requested intelligence support for at least 100 days, possibly through September. On March 3, Trump posted on Truth Social that wars can be fought "forever." Senator Mark Warner told reporters: "We have seen the goals for this operation change four or five times. I'm not sure which of those goals, if met, means we're at an endgame."

That alone should concern you regardless of where you stand politically. The military performance has been devastating on paper, CENTCOM says Iran's missile launch capacity is down 90 percent from day one. The CSIS estimated the first 100 hours of operations alone cost $3.7 billion, most of it unbudgeted.

Experts at the Brookings Institution have nonetheless warned this will be "more complicated than the White House may have hoped." Air campaigns have never delivered regime change in modern history. Iran is a country of 90 million people. And even from a position of severe military weakness, Iran is executing a deliberate strategy: not winning, but making this economically painful for everyone in the region for as long as possible.

But there is a specific chain of events, already in motion, that connects this war directly to the infrastructure, energy, capital, and talent the AI boom depends on. This article walks through each one, and then maps out the three scenarios that determine whether the AI industry gets a reality check, a slow squeeze, or something much worse.

Everyone knows about the oil risk. Most people do not understand the full picture. The Strait of Hormuz carries roughly 20 million barrels of oil per day, 20 percent of global petroleum consumption, according to the US Energy Information Administration via Al Jazeera. Qatar and UAE's LNG exports also pass through the same six-kilometre-wide bottleneck, making it the single most consequential shipping chokepoint on earth.

Tanker traffic has dropped to near zero. Over 150 ships are anchored outside the strait, unable to pass. Protection and indemnity insurance was withdrawn for vessels attempting transit, making the economic risk too high for ship owners. Maersk, CMA CGM, and Hapag-Lloyd have suspended transits. Iraq was forced to shut down operations at the Rumaila oil field after running out of storage capacity with nowhere to send the product.

The bypass pipelines that exist can handle approximately 2.6 million barrels per day, according to the Columbia University Energy Policy Center. The strait carries 18–19.5 million barrels per day. The bypass covers roughly 14 percent of that. The rest has no alternative route. Oil has already recorded its steepest weekly gains since 2020, with Brent trading around $90 a barrel. Traders and analysts are warning that a prolonged closure pushes it past $100.

Higher oil prices should benefit oil producers. Normally, yes. Not when you cannot load the tankers. Iraq cut production because it ran out of storage. Saudi Arabia's largest domestic refinery at Ras Tanura was shut down after a drone strike caused a fire. QatarEnergy declared force majeure on LNG shipments after Iranian drones struck the Ras Laffan complex, halting production entirely. Qatar produces roughly 20 percent of global LNG.

The Columbia Energy Policy Center identified the deeper structural problem. An oil price spike jeopardises the foundations of the Gulf's non-oil diversification, now far more important to these states' long-term plans than the oil price number itself. The IMF described the overall situation as "highly fluid" and warned that the full economic impact depends entirely on how long this lasts and how far the damage spreads. The Gulf's post-oil future, AI infrastructure, tourism, sovereign wealth investment, is funded by oil revenue. When that revenue stops flowing, the diversification budget gets squeezed. And AI was explicitly embedded in that diversification plan to the tune of trillions of dollars.

Dubai, Abu Dhabi, and Doha spent decades and hundreds of billions of dollars building a brand built on one thing: stability. In one week, that brand fractured. Iranian missiles and drones have hit targets across the UAE, Qatar, Bahrain, Kuwait, and Saudi Arabia. Airports across the region have closed or restricted operations. More than 11,000 flights across ten countries were cancelled in the first days of the conflict. Emirates and Qatar Airways suspended operations entirely. Short-term rental cancellations in the UAE doubled overnight.

The Atlantic Council described this as the first conflict since World War II to directly impact cities and facilities that serve as hubs of the global economy. Dubai handles 15 percent of the world's gold trade. European aviation safety authorities have extended guidance to avoid the entire region's airspace. These are not peripheral markets, they are globally systemic nodes, and they are now under fire.

The Gulf states are expat economies. The UAE's population is approximately 88 percent foreign nationals. Qatar is similar. Dubai's value proposition to international business, talent, and capital was built on "come here, it's safe, it's stable, it's neutral ground." That proposition has evaporated.

Twenty thousand Americans left the region in the first days of conflict. The UK, Australia, and Canada ran evacuation charter flights. High-net-worth individuals drove ten hours from Dubai to Riyadh to access functioning airports, with private jet prices reportedly hitting $350,000. In Qatar, roughly one-fifth of bank deposits belong to non-residents. In the UAE, the figure is around one-tenth. If those deposits start moving, the financial stability problem compounds everything else.

This is the pressure point almost nobody is discussing in relation to AI's future. Israel hosts over 400 R&D centres for Nvidia, Intel, Microsoft, Google, and Apple. Israeli tech represents 17 percent of GDP and 57 percent of exports. AI companies account for roughly 30 percent of the Israeli tech ecosystem but nearly half of all investment rounds.

The Gaza conflict from October 2023 showed the pattern clearly: between 15 and 20 percent of tech employees were called up as reservists. Some companies lost 30 percent of their workforce. More than 8,000 tech workers left the country in the first ten months. Startup funding froze. This conflict is larger. Israel has called up 20,000 additional reservists on top of the 50,000 already on duty. Nvidia's planned multi-billion-dollar campus in northern Israel now has a much more uncertain timeline. Every month that Israeli AI talent spends on war readiness is a month of innovation the global AI ecosystem does not get back.

In 2025, the White House announced over $5 trillion in investment commitments from Gulf states, with AI explicitly embedded in the deals. The UAE committed a $1.4 trillion ten-year investment framework. Saudi Arabia pledged $600 billion over four years. Qatar put up $1.2 trillion in "economic exchanges." $20 billion in Saudi-funded AI data centres on US soil were part of those commitments. Independent reporting at the time was already cautious. Reuters' own tally of specifically announced deals was materially lower than the headline totals, and analysts have stressed that announced investment commitments are not guaranteed to be fulfilled. But even to the extent these commitments were genuine, the math has now fundamentally changed.

The New Arab reported that Gulf officials are evaluating force majeure clauses and reviewing overseas investment commitments as war-driven financial strains accumulate. These sovereign wealth funds collectively hold around $5 trillion in reserves, the money exists. The question is whether the political will and financial logic to deploy it outward survives a prolonged war in which domestic reconstruction and defence spending are now competing priorities.

To understand why all six pressure points matter specifically to AI, you need to understand how this industry is actually being funded right now. It is not operating on normal economics.

Bloomberg published an analysis describing the modern AI cycle as an increasingly interconnected web of circular deals. Here is how it works. Microsoft invests billions in OpenAI. OpenAI uses that money to rent computing power on Microsoft Azure. That spending shows up as Azure revenue, which justifies more infrastructure investment, which creates more capacity for OpenAI to rent. Nvidia sells chips to all of them, then invests in OpenAI, which buys more Nvidia chips. Amazon invests in Anthropic, which rents AWS compute, which shows up as AWS revenue.

A handful of companies, Microsoft, Nvidia, Amazon, Meta, Google, OpenAI, Anthropic are simultaneously suppliers, customers, investors, and validators of each other. The demand signal is circular. It is not anchored to the broader market.

The scale is hard to internalise. Hyperscalers collectively expect to invest around $650 billion in AI infrastructure in 2026 alone. McKinsey estimates AI-ready data centres could require $5.2 trillion in capital expenditure by 2030. OpenAI raised $110 billion at a valuation of approximately $840 billion, despite projecting losses through at least 2029. Anthropic reached a $380 billion valuation. These are not profitable businesses valued on earnings. They are bets on a future that requires the investment machine to never stop.

And increasingly, this is not funded by free cash flow. The tech sector issued over $400 billion in bonds in 2025, and Barclays has forecast even higher corporate bond issuance tied to AI buildout in 2026. Private credit, infrastructure funds, insurance company balance sheets, and retail investors through REITs are all flowing into data centre financing. Morgan Stanley estimates a $1.5 trillion financing gap must be filled by external capital for the data centre buildout through 2028 alone.

This machine requires three things to keep functioning: affordable energy, available capital, and confidence. The war attacks all three simultaneously.

Energy is the most direct hit. AI data centres were already the single biggest bottleneck in the infrastructure buildout, a large share of planned 2026 projects were facing delays due to power constraints before this conflict started. Add an oil shock. Add Qatar's LNG production halted. Add natural gas prices spiking. New data centres become harder to build, harder to power, and harder to justify financially.

Capital is next. The war-fuelled oil price surge has already pushed back expected Federal Reserve rate cuts and increased inflation anxiety. When rates stay higher longer, equity valuations that depend on long-dated growth expectations compress. Every data centre project funded by private credit or bond issuance just got more expensive. Every AI startup burning through unprofitable growth just saw its runway get shorter.

Confidence is the most fragile variable. The AI boom is sustained by a collective belief that the investment will pay off eventually. That belief is already under independent scrutiny before this war: 74 percent of companies reportedly saw no tangible value from AI initiatives, and failure rates for enterprise AI projects run as high as 85 percent. Add a geopolitical shock that simultaneously disrupts Gulf investment flows, raises energy costs, and tightens credit and the circular loop that was making everything look healthy starts to show the cracks underneath.

The Cooling scenario is the most likely if the war resolves on a short timeline, the Federal Reserve's Christopher Waller argued in early March that an oil shock resolving within weeks is unlikely to have a persistent inflation impact, and analysts at major institutions have broadly characterised a short disruption as producing modest, temporary growth effects.

The Squeeze is where the trajectory is currently pointing if Hormuz remains commercially non-functional into April. The Pop requires multiple dominoes falling simultaneously but Rest of World has documented that several are already moving, including direct strike damage to Amazon Web Services facilities in the UAE and Bahrain.

One detail from Rest of World that crystallises the physical vulnerability: seventeen submarine cables pass through the Red Sea, carrying the majority of data traffic between Europe, Asia, and Africa. The Houthi resumption of Red Sea attacks has closed that corridor at the same time Hormuz is blocked. Both global data chokepoints are simultaneously inaccessible and that has never happened before. As Doug Madory, director of internet analysis at the network intelligence firm Kentik, told Rest of World: "Closing both choke points simultaneously would be a globally disruptive event." Repair ships cannot safely reach either passage.

The decisive variables are specific and measurable. Hormuz shipping: does commercial traffic return to anything close to normal, or do insurance premiums stay prohibitive? This is the single most important variable in the entire analysis. Qatar LNG: does production restart on a credible timeline, or does force majeure extend into April? Every week of closure adds compounding inflationary pressure across Europe and Asia.

Gulf investment reviews: do the evaluations of overseas financial commitments stay informal, or do they become formal policy decisions with specific AI-related deals suspended? Federal Reserve signals: does the energy-driven inflation prompt a shift in rate path expectations? Any signal that cuts are delayed further compresses AI valuations directly.

Private credit stress: are there signs of disorderly deleveraging in data centre financing markets? Reuters has already reported credit markets "dented" by war-related concerns about AI debt levels, with stress signals appearing in private credit vehicles including withdrawal limits being triggered at major funds. That is the canary in the coal mine for the Pop scenario.

The answers to those questions determine which scenario plays out.

If they lean positive, the US–Iran war gives AI a reality check, not a crash. If they lean negative together, the conversation about an Iran war AI bubble burst stops being theoretical. And it will happen not because AI stopped working. It will happen because the financing machine that was funding the pace of progress was more fragile than anyone in the industry wanted to admit.

For enterprise AI teams, this is not an abstract market question. The models, the compute, the capital, and the talent behind every AI deployment your organisation depends on are connected to the same infrastructure under stress right now. B

Bitbiased covers the AI stories that actually matter, before the mainstream catches up. Subscribe to stay ahead →