Iran’s Revolutionary Guards just named 18 US tech firms as military targets. The age of the civilian data centre is over.

At 8pm Tehran time on Tuesday, a new kind of front line was drawn, not through desert terrain or along a disputed border, but through the server farms, cloud regions, and corporate campuses of America’s largest technology companies. The Islamic Revolutionary Guard Corps published a statement on its official Sepah News channel naming 18 US firms, from Apple and Microsoft to Nvidia and Palantir, as “legitimate targets” in retaliation for what it described as their role in enabling American and Israeli assassination operations inside Iran.

The list reads like a roll call of the Nasdaq’s most valuable constituents. Microsoft, Apple, Alphabet, Meta, Amazon, Nvidia, Intel, Cisco, Oracle, Dell, HP, IBM, JPMorgan Chase, Tesla, General Electric, Boeing, and Palantir all appear alongside Spire Solutions and G42, the Abu Dhabi-based AI firm that has become a linchpin of the Gulf’s artificial intelligence ambitions. The IRGC gave employees at these companies across the Middle East an immediate evacuation warning, urging anyone within one kilometre of their facilities to leave.

The threat is extraordinary in its specificity. Rather than targeting military installations or government buildings, the IRGC has identified private-sector technology infrastructure as the mechanism through which, it alleges, the United States has been locating and killing senior Iranian officials. The statement declared that American ICT and AI companies are “the key element in designing and tracking terror targets,” and that “for every assassination and terrorist act in Iran, one facility or unit belonging to these companies will face destruction.”

The accusation is not made from thin air. Since Operation Epic Fury began on 28 February, the United States has struck more than 10,000 targets inside Iran, according to US Central Command. The Israeli Defence Forces claimed to have killed 40 senior commanders in a single operation that the IDF described as possible only because of military intelligence capabilities. Iran’s supreme leader, Ali Khamenei, was killed in an Israeli air strike on his compound that same day. Defence minister Aziz Nasirzadeh and IRGC commander Mohammad Pakpour followed. Dozens of high-ranking political and military figures, along with members of their families, have been killed in what Tehran describes as a sustained campaign of US-Israeli aggression.

The 💜 of EU tech

The latest rumblings from the EU tech scene, a story from our wise ol' founder Boris, and some questionable AI art. It's free, every week, in your inbox. Sign up now!

It is the role of artificial intelligence in enabling this campaign that has prompted the IRGC’s pivot toward commercial technology infrastructure as a theatre of war. Bloomberg reported in late March that Palantir’s chief technology officer described the Iran conflict as the first major war driven by AI, with advanced tools processing vast data sets to accelerate targeting decisions. The US military has confirmed using AI for drone navigation, intelligence analysis, and what it calls “target selection tools,” though it maintains a human remains in the decision loop. Nature published an editorial calling for a moratorium on AI in warfare until international law catches up.

The IRGC’s logic, however strained, runs as follows: if American cloud computing, artificial intelligence, and surveillance platforms are providing the infrastructure that makes precision strikes possible, then the companies operating that infrastructure are combatants. It is a framing that international humanitarian law does not support in any straightforward way, but the distinction may matter less than the operational reality. These companies have physical presences across the Gulf states, and those presences are now, by the IRGC’s declaration, in the crosshairs.

The exposure is enormous. Microsoft has committed $15 billion to expanding its operations in the UAE by 2029. Amazon has pledged $5 billion to an AI hub in Riyadh. Oracle, Cisco, and Nvidia announced a partnership with OpenAI to build an AI campus in the UAE. Google and Amazon Web Services are constructing dedicated cloud regions in Saudi Arabia scheduled to launch this year. According to analysts at TD Cowen, hyperscaler capital expenditure is forecast to exceed $600 billion in 2026, with roughly 75 per cent tied to AI infrastructure. A substantial portion of that money is flowing into the very region the IRGC is now threatening.

The timing underscores a tension that predates this conflict but has been dramatically sharpened by it. For years, US technology companies have been building massive data centre infrastructure in the Middle East, drawn by sovereign wealth capital, favourable energy costs, and proximity to growing markets in South Asia and Africa. Oracle alone has committed an estimated $156 billion in capital spending to its AI infrastructure buildout. These investments were made on the assumption that the Gulf states would remain stable, business-friendly environments. That assumption now looks fragile.

The IRGC’s threat is not purely rhetorical. Iran has already launched drone and missile strikes against targets in the region since the war began on 28 February, firing more than 500 ballistic and naval missiles and nearly 2,000 drones, according to Iran’s Fars News Agency. Roughly 60 per cent of those launches were aimed at US targets in the region. The Intercept reported that data centres in the UAE and Bahrain have already come under deliberate attack for the first time in military history, disrupting critical cloud infrastructure.

For the 18 named companies, the calculus is grim. Evacuating employees from Gulf offices is manageable. Relocating or hardening billions of dollars of physical infrastructure is not. And the reputational dimension cuts both ways: companies that are seen as too closely linked to military operations risk backlash in other markets, while companies that distance themselves from the US government risk losing defence contracts that have become a significant and growing revenue stream during the AI boom.

The inclusion of JPMorgan Chase, Tesla, General Electric, and Boeing on the list suggests the IRGC’s targeting framework extends beyond technology into broader US economic infrastructure. Boeing’s military division supplies fighter aircraft and munitions. GE manufactures jet engines used in military platforms. Tesla’s presence is harder to explain through a direct military logic, but the IRGC appears to be drawing its circle wide enough to create maximum economic pressure.

G42, the sole non-American company on the list, is perhaps the most telling inclusion. The Abu Dhabi firm has positioned itself as the Gulf’s flagship AI company, signing partnerships with Microsoft, OpenAI, and Cerebras while navigating US concerns about its historical ties to Chinese technology firms. Its inclusion signals that the IRGC views the Gulf states’ AI ambitions themselves, not merely the American companies servicing them, as part of the threat.

What happens next depends on whether the IRGC follows through and, if so, what form the attacks take. Cyberattacks on corporate infrastructure are one possibility and arguably the more likely initial vector, given Iran’s well-documented capabilities in that domain. Physical strikes against data centres or office buildings in the Gulf would represent a dramatic escalation and risk drawing the host nations, several of which have been attempting to stay neutral, directly into the conflict.

Either way, the IRGC’s statement marks a threshold. The regulation of AI and its use in sensitive contexts has until now been framed primarily as a governance challenge, a matter of compliance frameworks, risk assessments, and legislative timelines. What Iran’s threat makes plain is that AI infrastructure is now also a matter of physical security. The servers are no longer neutral. Whether they ever were is a question the next phase of this conflict may answer.