Trump’s Threats Against Iran’s Bridges and Power Plants Raise Serious War Crimes Questions

A preliminary point should be stated at once. The United States is not a party to Additional Protocol I. That does not place the conduct described here beyond legal scrutiny. The core rules relevant to this analysis, including distinction, proportionality, precautions in attack, and the protection of medical units and civilian objects, are widely treated as rules of customary international law binding on all parties to an armed conflict, even where the United States has objected to particular formulations or treaty language. The analysis that follows proceeds on that footing.

A second point is equally necessary. This article does not purport to adjudicate the conflict as a whole. It examines a defined body of reported U.S. conduct and threats: the reported strike on the Shajareh Tayyebeh girls’ school in Minab on 28 February 2026, the WHO verified pattern of attacks on health facilities in March, the reported bridge strike near Tehran on 2 April, and the public threats reported by Reuters and AP on 5 and 6 April to destroy Iranian bridges and power plants if Tehran did not reopen the Strait of Hormuz. The conduct of other parties raises separate legal questions. They are not addressed here.

The legal issue is narrower than much of the public commentary suggests, but more serious. The stronger case on the presently available record is not genocide. It is whether actual and threatened attacks on schools, health facilities, bridges, and essential infrastructure engage the prohibition on attacking civilian objects, the rules governing proportionality and precautions in attack, and the prohibition on threatening conduct that would itself be unlawful if carried out. Those are distinct legal disciplines. They are already engaged by the material now in the public domain.

The chronology is short and already troubling. Reuters reported that U.S. investigators believed U.S. forces were likely responsible for the 28 February strike on the Minab girls’ school, although the investigation had not then reached a final conclusion. Reuters later reported that outdated targeting data may have contributed to that strike. On 5 March, Reuters reported that the World Health Organization had verified 13 attacks on health sites in Iran. By mid March, that total had risen, and Reuters reported hospital evacuations and deaths among health workers. On 2 April, AP reported that a bridge near Tehran was struck during Nature Day celebrations, with Iran alleging eight civilians killed and ninety five injured. On 5 and 6 April, Reuters and AP reported threats by Donald Trump to destroy Iranian bridges and power plants if Tehran did not comply over Hormuz. Those incidents are sufficient to require legal examination rule by rule.

Civilian objects: the prohibition is basic, and the Minab strike is the test case

The first rule is elementary. Civilian objects may not be attacked as such. A military objective is not whatever a belligerent chooses to describe in those terms after the fact. The orthodox test requires both limbs: the object must make an effective contribution to military action by its nature, location, purpose, or use, and its destruction, capture, or neutralisation must offer a definite military advantage in the circumstances ruling at the time. If those elements are not met, the object remains civilian and attack upon it is unlawful.

The Minab school places that rule under direct pressure. On the public record, it appeared to be what it said it was: a girls’ school. Reuters reported that U.S. military investigators considered U.S. forces likely responsible, but that the investigation had not concluded. That caveat matters. It does not remove the legal concern. If a girls’ school was struck because stale or defective target data was used, the issue is whether a plainly civilian educational facility was unlawfully treated as a military objective.

Doubt is not a licence to strike: the bridge case and the presumption of civilian use

The second issue concerns infrastructure that is civilian in character but may be said to have military utility. This is where weak analysis usually collapses into slogan. Either everything is treated as civilian and therefore immune, or everything useful is treated as dual use and therefore fair game. The law accepts neither shortcut. Where there is genuine doubt whether an object normally dedicated to civilian purposes is being used to make an effective contribution to military action, the attacker still has to justify the target in the circumstances ruling at the time.

The reported strike on the bridge near Tehran is the clearest example. AP reported civilian casualties on a public holiday. The obvious defence is dual use. Bridges can carry military traffic. That is not the end of the inquiry. It is the beginning. The operative question is whether, at the time of strike, on the intelligence available, the bridge had in fact become a lawful military objective, not whether bridges in the abstract may sometimes be struck. On the present public record, no targeting file has been disclosed, no intelligence basis has been produced, and no serious target specific explanation has been made public.

Medical facilities: WHO verification does not prove liability, but it destroys any pretence that the issue is speculative

Hospitals and medical units occupy a separate and specially protected category. That protection is not decorative. It can be lost only in narrow circumstances and ordinarily only after warning where the law requires it. Repeated strikes on health infrastructure are therefore not just more examples of civilian harm. They engage one of the clearest protected classes in the law of armed conflict.

Here the quality of the evidence matters. WHO verification is not the same thing as a final adjudicated finding. It does not assign criminal liability. It does not, by itself, determine whether a particular facility may have lost protection in a particular case. But it does establish something that cannot honestly be reduced to propaganda or battlefield confusion: repeated attacks on functioning health facilities were verified by the WHO in March 2026, and Reuters reported hospital evacuations and deaths among health workers.

Proportionality: even a lawful target is not lawfully attackable at any civilian cost

Proportionality is where loose political talk runs into hard legal structure. Even if a target is a lawful military objective, an attack is prohibited where the expected incidental civilian harm would be excessive in relation to the concrete and direct military advantage anticipated. The assessment is ex ante, judged on the information reasonably available at the time. Public information may be incomplete, but incomplete evidence does not dissolve the rule.

The bridge strike again provides the clearest illustration. Civilian casualties were reported during Nature Day celebrations. Even assuming, for present purposes, some military use, the legal question remains whether the expected civilian harm in those circumstances was excessive relative to the specific advantage sought. That question has not been answered in public. The same analysis becomes more severe in relation to power infrastructure. A power facility may, in some circumstances, support military systems. But if its destruction foreseeably disables hospitals, communications, sanitation, refrigeration, and basic civilian life over a wide area, the proportionality analysis becomes exacting, not relaxed.

Precautions in attack: stale targeting data is not an administrative footnote, it is a legal problem

The duty of precaution is often treated in public debate as mere procedure. It is nothing of the sort. The obligation to do everything feasible to verify the target, to choose means and methods that reduce civilian harm, and to cancel or suspend where necessary is a substantive legal discipline. Careless targeting is one of the ways unlawfulness occurs.

Reuters’ reporting that outdated targeting data may have contributed to the Minab strike is therefore not a side issue. If accurate, it goes directly to precautions. It suggests a failure in the verification process itself. The investigation was not complete, and that caveat must remain. But if the target package was built on stale data and a girls’ school was struck as a result, the question is whether the targeting discipline required by the law of armed conflict failed at the point where it mattered most.

Infrastructure indispensable to civilian life: broad threats against the power system are legally indefensible on their face

Objects indispensable to civilian survival sit at the centre of the law’s attempt to prevent warfare from collapsing into engineered social ruin. Modern civilian life is system dependent. Electricity underpins healthcare, water treatment, sanitation, communications, refrigeration, and basic habitability. A threat directed at the national power system is therefore not merely a threat against machinery. It is a threat against the conditions of civilian life.

AP reported a threat to destroy all Iranian power plants. That reported wording is load bearing and should be treated carefully. Unless and until a primary transcript is produced, the analysis must remain tied to the terms as reported. But on those terms, the legal problem is acute. A facility by facility analysis may in some cases justify attack on individual power assets. A blanket threat against all power plants cannot be defended on the basis that some members of that class might, somewhere, under certain conditions, be lawful military objectives. The law requires target specificity. A category wide threat against an essential civilian system advertises the opposite.

Threats of unlawful force: words do legal work

The ICRC President was right to say that the rules of war must be respected in words and action. That was a statement of law. A state cannot lawfully threaten conduct that would itself be unlawful if carried out. That principle does not depend on whether the threat is later executed. It is enough that the threat is publicly made and that the conduct threatened would breach the applicable rules if implemented.

That is why attempts to dismiss the reported Trump threats as mere rhetoric do not succeed. Political rhetoric is not a legal safe harbour. If the reported ultimatum was that Hormuz must be reopened or Iran’s bridges and power plants would be destroyed, then the legal question is straightforward: would the threatened conduct be lawful if carried out in those terms? On the present analysis, there is a serious basis to say that it would not.

Collective punishment: the concern is real, but the classification should be controlled

There is a temptation to go further here and declare collective punishment. That temptation should be resisted unless the evidence justifies it. Geneva Convention IV prohibits collective penalties. The concept is real and serious. It is not a catch all label for every broad or coercive measure. It requires careful application.

The concern should nevertheless be stated plainly. A reported threat framed as reopen the strait or lose the bridges and power plants carries the logic of pressure on an entire society through essential systems in order to compel political compliance. That is proximate to the logic the prohibition on collective punishment exists to prevent. Fuller evidence would be required before any concluded classification is made.

Genocide: not the strongest present claim

The Genocide Convention requires proof of specific intent to destroy, in whole or in part, a protected group as such. The public material presently available does not safely establish that threshold. It is more consistent with coercion, devastation, infrastructure pressure, and possible war crimes than with the special intent required for genocide.

On the present record, war crimes analysis is the stronger line. The law should not be weakened by attaching to the facts a category that the evidence does not yet safely bear.

The defence case is real, but it does not neutralise the issue

A serious analysis must state the defence case in its strongest form. It would say that bridges, energy nodes, and transport infrastructure are dual use and integrated into military logistics. It would say no outside analyst has the targeting file. It would say Reuters, AP, and WHO reporting are not substitutes for adjudication. All of that is true. None of it disposes of the legal problem.

Those points may narrow some allegations. They do not explain away a girls’ school reportedly struck by forces under investigation. They do not erase WHO verification of repeated attacks on health facilities. They do not solve the problem created by a public threat framed in sweeping, category wide terms against an entire class of civilian dependent infrastructure. The public record is incomplete. It is not inert.

Conclusion

The present record does not establish criminal liability. It raises serious, concrete, and immediate questions under the law of armed conflict in relation to a reported school strike, repeated verified attacks on health infrastructure, a bridge strike with reported civilian casualties, and public threats to destroy bridges and power plants in terms so broad that they are difficult to reconcile with the most basic rules of targeting law.

On the presently available public evidence, investigation is warranted; explanation is overdue; and if the conduct reported is confirmed, there is a serious basis for concluding that core obligations of the law of armed conflict were breached. The strongest present case is not genocide. It is that attacks and threats directed at civilian objects and essential infrastructure have moved from allegation into the territory of structured legal concern.