2026 is already a landmark year for US military prowess. In early January, the US seized Nicolás Maduro in a stunning operation – codenamed Operation Absolute Resolve – with no American fatalities. American intelligence, reportedly aided by AI systems including Anthropic’s Claude, offered a clear window into the Venezuelan president’s movements, giving the Trump administration the freedom to choose when to strike.[1] After suppressing Venezuelan air defences through a mix of electronic warfare and kinetic strikes, helicopters carried US special forces into the heart of Caracas to capture Maduro. The US Chairman of the Joint Chiefs of Staff, General Dan Caine, hailed it as “an audacious operation that only the United States could do”.[2] Trump declared it “one of the most stunning, effective, and powerful displays of American military might and competence in American history”.

Rapid success in South America emboldened the White House. On 28 February, US and Israeli forces launched a devastating wave of attacks against Iran, killing Supreme Leader Ali Khamenei and several senior military commanders, including the head of the Islamic Revolutionary Guard Corps (IRGC), in a matter of hours. AI systems developed by companies such as Palantir and Anduril – “the darlings” of the newly renamed Department of War – again played a key role, with Palantir’s Maven system, powered by Claude, reportedly sifting through vast quantities of intelligence data to condense what would previously have taken days of targeting work into minutes.[3] Never before has it been possible to decapitate an adversary’s senior figures at the opening of hostilities with such speed.

In the weeks that followed, the US and Israel systematically hunted down further layers of Iran’s command structure, including intelligence minister Esmail Khatib, security chief Ali Larijani, and the head of the paramilitary Basij force, Gholamreza Soleimani. The sustained campaign, supported by the intelligence advantages that AI targeting systems provide, stands in stark contrast to Russia’s failed attempt to assassinate Ukrainian President Volodymyr Zelenskyy in February 2022, lending some weight to Caine and Trump’s claim that only the US can pull off operations of this kind.

The wider US-Israeli military campaign first targeted Iran’s air defences with expensive stand-off weapons. Having established air supremacy, they have been able to strike the rest of Iran’s military assets at will using relatively cheaper munitions. Iran’s navy and air force have been crippled and its surviving leadership has been forced into hiding. As US Defense Secretary Pete Hegseth put it on 8 March, “this is not a fair fight, and that’s on purpose. Our capabilities are overwhelming compared to what Iran’s are”.[4] Leaving aside the bombast, Iran is no military pygmy, but this only underscores the tactical skill required to degrade its military capabilities in such short order.

Yet for all of these tech-enabled military marvels, the US is yet to turn overwhelming military force into strategic victory in Iran (to the extent that core strategic objectives can be discerned.) This war is reinforcing an enduring lesson: the adversary always gets a vote. More broadly, the Trump administration’s growing willingness to use force – and the technological advances that make it so tempting – carry significant costs, from munitions stocks to moral authority, that are beginning to surface.

Trump is not the first, nor will he be the last, leader to launch a war on the assumption that it would be quick. The 1967 Six Day War, the Falklands War (74 days), and the 1991 Gulf War (42 days) are exceptions to a general rule of protraction.

In his study of this phenomenon, Iskander Rehman notes that “throughout history, short-war thinking has all too often run aground on the jagged shoals of political reality”.[7] Few of the belligerents in 1914 would have predicted that the war would be as lengthy and costly as it transpired. Years later, Hitler invaded the Soviet Union on the assumption that “we have only to kick in the door and the whole rotten structure will come crashing down”. Similarly, and more recently, the Russian military anticipated Kyiv would fall within days or weeks of its full-scale invasion in 2022. All were mistaken.

During the Second World War, the Germans underestimated the Soviets’ ability to shift their industry east and recruit more men. In Ukraine in 2022, the Kremlin failed to appreciate the wells of Ukrainian nationalism and the skill of Volodymyr Zelenskyy in drawing on them. The doyen of Strategic Studies, Sir Lawrence Freedman, writes that “despite this long history of protracted conflict, military strategists continue to shape their thinking around short wars, in which all is supposed to be decided in the first days, or even hours, of combat”.[8]

The short war fallacy often stems from underestimating an adversary’s resolve and capabilities. This appears to be the case today – with Trump recently admitting that Iran’s response was “the biggest surprise I had of this whole thing”.[9] Following the assassination of its political leadership, Iran not only fired missiles at US and Israeli targets but also cities across the Gulf. It has since expanded its attacks to shipping and energy infrastructure. Despite the degradation of its missile and drone capabilities from the US-Israeli air campaign, Iran is maintaining a persistent rate of fire that is sufficient to prevent the resumption of normal economic life across the region – with knock-on effects for markets across the world. The Iranians know they are outgunned by America’s technological wizardry, but they can fight asymmetrically and take advantage of their geostrategic location. Despite the pounding it has taken, Iran has ironically gained leverage by constricting one of the global economy’s arteries. Rather than weaken their stranglehold over the Strait of Hormuz, the war has only proven that Iran has effective control over those waters. Even now, in the midst of the war, Iran is able to make demands for toll fees for safe passage.

The two-week ceasefire announced earlier this week has already been tested.  To the extent the ceasefire offers a pause, there is still no guarantee that Iran will not resume hostilities once the two week period has expired, given its hold over one of the world’s chokepoints. Trump has already declared victory on several occasions; on 11 March, for example, he told a crowd: “let me tell you, we’ve won…in the first hour, it was over”.[10] Yet the precarious state of global energy markets suggests otherwise. As Prussian general Helmuth von Moltke the Elder observed, no plan survives contact with the enemy. Perhaps Mike Tyson put it better: “everyone has a plan until they get punched in the mouth”.

Prior to the ceasefire, the Trump administration was reportedly considering deploying ground forces.  That option appears to remain on the table, depending on the durability of the ceasefire.  Judging by the assets despatched to the region, including elements of the 82^nd Airborne and a Marine Expeditionary Unit, any action would likely be limited to short-term raiding – keeping with the mantra of short, sharp, tech-facilitated strike. For context, the regime change operation in Iraq in 2003 involved over 250,000 US troops, five times the number stationed in the region today.

The seeming success of ‘regime alteration’ in Venezuela appears to have convinced Trump that the model could be emulated. “What we did in Venezuela”, he told the New York Times a few days into Operation Epic Fury, “is the perfect, the perfect scenario”.[5] Yet it must also be acknowledged that the Maduro raid involved a dose of good fortune; one US pilot was shot several times but, incredibly, still managed to guide his helicopter to the Venezuelan president’s compound. Had the pilot died – resulting in the loss of a helicopter laden with commandos – the narrative could well have been different.

Beyond chance and contingency, the idea that a model of regime alteration can be inferred from one case in South America and transferred to the Middle East is gravely misconceived. Even leaving aside the seismic cultural differences between Venezuela and Iran, the context is wholly dissimilar. Unlike Caracas, where the Americans dealt with Maduro’s vice president, Delcy Rodriguez, there is no clear equivalent in Tehran. Anti-Americanism is baked firmly into the Islamic Republic’s ideology and there are hundreds of thousands, if not millions, of individuals invested in the regime’s survival.

The Iranian regime is also increasingly decentralised, adopting a ‘mosaic defense’ strategy, which explains its success in retaliating so swiftly after the initial strikes.[6] Indeed, Israel already tested the decapitation approach on Iran less than a year ago; the June 2025 strikes killed the commander of the IRGC, the armed forces chief of staff, and several nuclear scientists. The regime reconstituted its leadership within weeks. Moreover, Iran’s geostrategic position – dominating the Strait of Hormuz through which one-fifth of the world’s oil and a significant share of its liquefied natural gas flow – offers its remaining rulers the kind of leverage that the Venezuelans could only dream of.

While the ongoing operation is displaying the formidable power of the US military, it is important to appreciate the trade-offs in using so many exquisite munitions in this theatre. The US is racing through its stocks of stand-off weapons and air defence systems, both of which would be essential in a Taiwan contingency. Thus far, it has fired approximately 850 Tomahawk land attack missiles, more than the total expended during the 2003 Iraq War. Its stocks are estimated to be fewer than 3,000, while new deliveries are around 190 per year.[11] The war is also depleting air defence stocks, notably the high-end THAAD and Patriot systems. Recent analysis from the Royal United Services Institute suggests the US fired 40% of its total THAAD inventory in the first 16 days of the war.[12] Moreover, on some occasions, where air defences have been saturated, these expensive interceptors (a PAC 3 costs around $4.1m) have been used against $30,000 Iranian Shahed drones. The US and its partners have deep pockets, but they are on the wrong side of the cost equation.

Military force has a finite character that, once employed, denudes its potential elsewhere. In his study of Roman grand strategy, the strategist Edward Luttwak observed that military “power born of potential force is not expended when used, nor is it a finite quantity. Force, on the other hand, is just that: if directed to one purpose, it cannot simultaneously be directed at another, and if used, it is ipso facto consumed.”[13] A similar dynamic is visible today, as the drain on resources in the Middle East is already having consequences elsewhere. The US reportedly pulled dozens of THAAD air defence missiles out of South Korea in March – undermining deterrence vis-à-vis North Korea and China – and is apparently pressing some European allies to send their Patriot systems to the Middle East.[14] In short, military operations against Iran should not be viewed in isolation; the longer this war continues, the weaker the US and its allies’ air defences get elsewhere.  By extension, the US military deterrence in East Asia and Europe softens.

The Trump administration’s increasing willingness to use unilateral force is matched only by its disdain for the laws and norms associated with the employment of such force. It justifies its actions in a transactional, even brutish, manner. Earlier this year, Treasury Secretary Scott Bessent defended Trump’s threats to annex Greenland: “We are the strongest country in the world. Europeans project weakness. The US projects strength.”^[15] More recently, there was never any question of the administration going through the UN to endorse the Iran operation. This marks a rupture with past practice. For decades, American leaders “cloaked U.S. power in the language of law, legitimacy, or universal liberal values”.[16] At times, they fell short of these lofty standards, notably over Iraq in 2003, but that very hypocrisy served as a constraint. Washington incurred reputational costs with weaker states if it deviated from its own rhetoric.

The current administration is voluntarily surrendering any claim to a special position in the international system; the US is just another great power ruthlessly pursuing its interests.

This abandonment carries wider strategic costs. The administration refrains from labelling its war in Iran as a ‘war’, preferring the term ‘major combat operations’ – language that conveniently sidesteps the need to seek congressional approval. Yet this is precisely the kind of euphemism Washington has long condemned in others. Russia’s illegal war in Ukraine is similarly dressed up as a ‘special military operation’. By adopting the same playbook, the US forfeits the moral authority to call out such language when its adversaries use it.

The growing preference for assassinating the political and military elites of hostile states represents a break with a decades-old tradition of restraint – if imperfectly observed. The US is no stranger to targeted assassinations; President Obama routinely used drone strikes against terrorist groups. There is, however, a longstanding taboo when it comes to states. Since the Ford administration, successive presidents have maintained an executive order prohibiting the assassination of foreign leaders.[17] The convention was never absolute (Muammar Gaddafi and Saddam Hussein were targeted during active military operations), but it broadly endured as a peacetime norm. Trump’s assassination of Qasem Soleimani, the head of the IRGC’s Quds Force, in 2020 signalled a departure. What is happening now, however, represents a qualitative escalation. Instead of targeting one individual military commander, the US and Israel are methodically eliminating a hostile state’s political and military command structure. The normalisation of decapitation as an instrument of statecraft should unsettle even those who welcome the weakening of Iran.

Finally, the fusion of AI and precision weapons, hailed in the opening days of the campaign, raises its own ethical questions. The technology that shrinks the ‘kill chain’ from days to minutes is remarkable, but it also increases the speed with which the military can make catastrophic errors. The same AI systems credited with accelerating the targeting cycle may turn out to have had a role in one of the war’s darkest episodes.[18] On 28 February, a US strike destroyed a girls’ school adjacent to an IRGC base in southern Iran, resulting in the deaths of over 160 people, most of whom were children. The Pentagon has opened an investigation into whether outdated intelligence (bad data from human error) or AI-assisted targeting was at fault.

Any verdict is unlikely to be forthcoming. Secretary Hegseth cut the Civilian Protection Center of Excellence workforce by approximately 90% and slashed CENTCOM’s civilian casualty assessment team from 10 to one, as part of his focus on lethality.[19] Just as AI becomes wired into US military operations, the guardrails are coming off.

The Trump administration’s campaigns in Venezuela and Iran have demonstrated a level of tactical and technological proficiency that is the envy of military planners worldwide. Having fused precision weaponry and AI-aided targeting, the US military can offer the president options without historical parallel. Taken narrowly, Hegseth and Trump are right to assert that no other country could do (or has done) what the US has been able to do in recent months.

Iran, however, demonstrates that the adversary still has agency. Despite absorbing the most devastating opening salvo in modern military history, Tehran shows little sign of unilaterally conceding. Its leadership has been decimated, its air force and navy crippled, yet it retains enough asymmetric capability to hold the global economy hostage through the Strait of Hormuz. The campaign, meanwhile, has consumed munitions at a rate that weakens American deterrence in other theatres, notably East Asia. More broadly, the Trump administration’s readiness to bypass legal and ethical constraints is undermining the very order which the country has so long upheld.

Trump and his commanders have proved they can break things faster than any military in history. They have not yet shown they know what to do with the pieces.

[1] Dave Lawler & Maria Curi, ‘Pentagon used Anthropic’s Claude during Maduro raid’, Axios, 14 February 2026, https://www.axios.com/2026/02/13/anthropic-claude-maduro-raid-pentagon.

[2] The Department of War, ‘Trump announces U.S. military’s capture of Maduro’, 3 January 2026, https://www.war.gov/News/News-Stories/Article/Article/4370431/trump-announces-us-militarys-capture-of-maduro/.

[3] John Bew, ‘The age of invasion’, The New Statesman, 7 January 2026; Michael Shepard, Rebecca Torrence, and Edward Ludlow, ‘Palantir CTO sees Iran War as first major AI-driven conflict’, Bloomberg, 25 March 2026, https://www.bloomberg.com/news/articles/2026-03-24/palantir-cto-sees-iran-war-as-first-major-conflict-driven-by-ai.

[4] Major Garrett et al, ‘On U.S. strikes against Iran, Pete Hegseth says, “this is only just the beginning”’, CBS News, 8 March 2026, https://www.cbsnews.com/news/hegseth-us-strikes-iran-trump-plans-60-minutes-transcript/.

[5] Zolan Kanno-Youngs, David E. Sanger and Tyler Pager, ‘Trump says war could last weeks and offers contradictory visions of new regime’, New York Times, 1 March 2026, https://www.nytimes.com/2026/03/01/us/politics/trump-iran-war-interview.html.

[6] Ashkan Hashemipour, ‘The IRGC’s way of war’, Engelsberg Ideas, 16 March 2026, https://engelsbergideas.com/notebook/the-irgcs-way-of-war/.

[7] Iskander Rehman, Planning for Protraction: A Historically Informed Approach to Great-power War and Sino-US Competition (London: International Institute for Strategic Studies, 2023), p. 20.

[8] Lawrence Freedman, ‘The age of forever wars: why military strategy no longer delivers victory’, Foreign Affairs (May/June 2025).

[9] Kristen Welker & Alexandra Marquez, ‘Trump says Iran is ready to negotiate a ceasefire but he’s not ready to make a deal’, NBC News, 15 March 2026, https://www.nbcnews.com/politics/donald-trump/iran-negotiate-ceasefire-deal-trump-kharg-hormuz-oil-rcna263474.

[10] Zachary B. Wolf et al, ‘Epic flurry: How Trump’s words on Iran have yo-yoed over three weeks of war’, CNN, 25 March 2026, https://edition.cnn.com/2026/03/25/politics/watch-how-trump-has-shifted-his-stance-over-three-weeks-on-almost-everything-iran.

[11] Mark F. Cancian and Chris H. Park, ‘The 850 Tomahawks launched in Operation Epic Fury is the most fired in a single campaign’, Center for Strategic & International Studies, 27 March 2026, https://www.csis.org/analysis/850-tomahawks-launched-operation-epic-fury-most-fired-single-campaign.

[12] Macdonald Amoah, Morgan Brazilian and Jahara Matisek, ‘Over 11,000 munitions in 16 Days of the Iran War’, Royal United Services Institute, 24 March 2026, https://www.rusi.org/explore-our-research/publications/commentary/over-11000-munitions-16-days-iran-war-command-reload-governs-endurance.

[13] Edward N. Luttwak, The Grand Strategy of the Roman Empire: From the First Century A.D. to the Third (Baltimore: Johns Hopkins University Press, 1976), p. 33.

[14] Kim Hyun-soo, ‘Relocation of THAAD interceptors to Middle East seen as imminent: sources’, Yonhap News Agency, 11 March 2026, https://en.yna.co.kr/view/AEN20260311010200315; Victor Jack and Jan Cienski, ‘Poland won’t divert Patriot air defense systems to Gulf’, Politico, 31 March 2026, https://www.politico.eu/article/poland-wont-divert-patriot-air-defense-systems-to-gulf/.

[15] Scott Bessent quoted in NBC News interview, 18 January 2026, https://www.nbcnews.com/politics/trump-administration/treasury-secretary-bessent-tariffs-national-emergency-greenland-eu-rcna254650

[16] Matias Spektor, ‘The world will come to miss western hypocrisy’, Foreign Affairs, 29 January 2026.

[17] Tim Naftali, ‘Trump opens the Pandora’s box of assassination’, The Atlantic, 3 March 2026, https://www.theatlantic.com/ideas/2026/03/khamenei-assassination-precedent/686215/.

[18] Katie Livingstone, ‘Deadly Iran school strike casts shadow over Pentagon’s AI targeting push’, The Military Times, 25 March 2026, https://www.militarytimes.com/news/your-military/2026/03/24/deadly-iran-school-strike-casts-shadow-over-pentagons-ai-targeting-push/.

[19] Paul McLeary and Jack Detsch, ‘Hegseth gutted offices that would have probed Iran school strike’, Politico, 10 March 2026, https://www.politico.com/news/2026/03/10/pentagon-iran-school-strike-civilian-casualties-00820780.