Were the Iran critical infrastructure strikes lawful under international law?

Legal analysis of the US-Iran critical infrastructure strikes has to begin with a practical question, not a slogan: when an air campaign reaches power plants, desalination facilities, gas infrastructure, schools, health facilities, and communications-linked systems, what must international law ask before anyone can call those strikes lawful or unlawful?

The provisional answer is uncomfortable but not especially complicated. Credible violation theories exist under both the law governing resort to force and the law governing conduct in hostilities. They concern authorization, imminence, distinction, proportionality, precautions, and the special protection of objects indispensable to civilian survival. But recognition of a plausible violation theory is not the same thing as enforcement. As of July 19, 2026, the conflict is ongoing, the public record is incomplete, classified targeting material is unavailable, and the most obvious international criminal forum faces serious jurisdictional barriers.

This article is an informational legal briefing for professionals monitoring international humanitarian law developments. It is not legal advice, and it does not assume that a court has already found facts that remain contested.

The reported pattern matters because infrastructure is not a legal category by itself

The 2026 Iran war moved quickly from a dispute over force protection and nuclear-related concerns into a broader campaign whose reported targets included energy, water, and civilian-linked facilities. Britannica’s conflict overview provides the basic chronology, including the outbreak of hostilities, the later memorandum of understanding, and the resumption of hostilities after the June 2026 breakdown of that arrangement.[1]

The legal significance does not come from the word “infrastructure” alone. A bridge, grid node, power plant, fuel depot, communications relay, school building, hospital wing, water system, or gas facility can sit in very different legal positions depending on use, location, expected military advantage, civilian dependence, feasible alternatives, and the attacker’s information at the time. The same physical object may also change character over time. That is why dual-use analysis is necessary. It is also why dual-use cannot be allowed to become a permission slip.

Public reporting has described strikes and threats involving civilian infrastructure, including a school strike and wider concern over attacks on systems supporting civilian life.[2] NPR’s coverage likewise treated the public debate as one about possible war crimes, not merely about policy wisdom or escalation management.[3] Those reports do not settle the legal question. They do show that the legal controversy is not speculative.

Impact figures should be handled with care. Airwars reported 1,443 civilian deaths, including 217 children, for the period from February 28 to March 23, while Iranian Red Crescent data reported 67,414 civilian sites struck, including 498 schools and 236 health facilities.[4][5] Those numbers were produced under active hostilities and access constraints. They are not a substitute for incident-level legal findings. They are, however, relevant to the scale of foreseeable civilian harm, the pressure on essential services, and the evidentiary questions any later investigation would ask.

The first legal problem is the resort to force

Before reaching targeting law, the analysis has to ask whether the use of force itself was lawful. That is the jus ad bellum question. The strongest public legal intervention on that issue is the Just Security open letter signed by 135 professors, which argued that the United States lacked a valid legal basis for the war because there was no UN Security Council authorization and no credible public showing that Iran posed an imminent armed attack justifying self-defense.[6]

That argument matters because infrastructure strikes do not become lawful at the strategic level simply because a military planner can describe their operational utility. If the resort to force itself violates the UN Charter framework, every subsequent use of force sits under that initial defect. That does not make the conduct rules irrelevant; jus ad bellum and jus in bello remain analytically distinct. A state can fight an unlawful war while still being bound to distinguish civilians from military objectives, take precautions, and avoid disproportionate attacks. But the absence of a lawful basis for force is not a background concern. It is a threshold legal problem.

The imminence point is especially important. International law does not require a defending state to absorb the first blow in every circumstance. It also does not permit a government to convert long-running hostility, generalized threat perception, or strategic anxiety into a standing license for preventive war. On the public record summarized by the Just Security signatories, the asserted legal basis had not been matched by evidence of an imminent armed attack of the kind that would support self-defense.[6]

That conclusion is necessarily public-record limited. Governments often claim that the decisive intelligence cannot be disclosed. Lawyers who have worked near targeting and national security processes know that this is sometimes true. They also know that secrecy cannot do all the work. If the public legal justification remains a formula while the military campaign reaches systems that sustain civilian life, the discipline expected of Article 2(4) and Article 51 analysis begins to erode.

Dual-use energy systems do not lose civilian protection wholesale

The conduct-of-hostilities question is narrower and more exacting. Additional Protocol I’s targeting framework, reflected in customary IHL debates and routinely used as an analytical baseline, asks whether an object is a military objective by nature, location, purpose, or use, and whether its destruction, capture, or neutralization offers a definite military advantage in the circumstances ruling at the time. The ICRC’s work on energy infrastructure emphasizes that attacks on such systems can generate reverberating effects far beyond the immediate blast: hospitals lose power, water pumps stop, sewage treatment fails, food storage deteriorates, and repair crews operate under continuing danger.[7]

That is the point often flattened in public descriptions of grid or fuel targets. A power plant may support command-and-control systems, radar, weapons production, or military logistics. It may also keep incubators running, dialysis machines operating, water treatment moving, and apartment blocks habitable in extreme weather. A serious legal analysis has to hold both facts in view at the same time.

The Yale Law Journal’s study of the rise of dual-use objects in war warns against an expansive reading that allows civilian objects to be swept into military-objective status because some military use can be identified somewhere in the system.[8] The danger is not theoretical. Modern societies route civilian and military functions through shared grids, ports, communications networks, transport systems, fuel supply chains, and data infrastructure. If every shared function makes the whole network targetable, the distinction principle loses much of its practical force.

The Lieber Institute’s analysis of attacks on power infrastructure reaches the same operationally sober point: power infrastructure is not categorically immune, but neither is it categorically targetable. The attacker must identify the specific military use and the concrete and direct military advantage expected from the attack, then assess expected incidental civilian harm, including reasonably foreseeable knock-on effects.[9]

The reported inclusion of desalination facilities and water-linked systems is therefore not just another item in a target list. Water infrastructure can fall within protections for objects indispensable to civilian survival, depending on use and context. The ICRC President’s warning that a “war on essential infrastructure is a war on civilians” was not a claim that every attack on infrastructure is automatically a war crime; it was a reminder that deliberate attacks on essential services and civilian infrastructure can amount to war crimes when the legal elements are met.[10]

Power, gas, and desalination targets require more than a systems-level rationale

A systems-level rationale is tempting in modern war. Disable the grid, disrupt fuel flows, strain communications, slow repairs, and the adversary’s military machine has to spend time improvising. But IHL does not ask whether attacking a broad system would be useful. It asks whether the specific object attacked is a military objective, whether the expected advantage is concrete and direct, and whether expected civilian harm is excessive in relation to that advantage.

C4Defence’s analysis of the US-Iran grid conflict describes the operational logic of energy targeting, including the military relevance that energy systems can have in a conflict involving air defense, command systems, and industrial capacity.[11] Baker Botts’ practitioner analysis of critical infrastructure and the law of war similarly treats these targets as legally fact-dependent rather than inherently prohibited or inherently lawful.[12] That is the right starting point. The problem begins when “dual-use” is treated as the end of the inquiry.

Consider a hypothetical example. If one substation directly and currently powers an integrated air-defense node, the military-objective argument may be stronger than it would be for a regional power plant serving mostly civilian neighborhoods and only attenuated military demand. Even then, the proportionality and precautions analysis would not disappear. The attacker would still have to consider timing, weapon choice, expected outages, hospital backup capacity, repair feasibility, warnings where effective, and whether a narrower attack could achieve the same concrete military advantage with less civilian harm.

The reported South Pars gas field strikes raise a related question. Energy production can support military operations and state revenue. But revenue impairment is a legally fraught theory if the asserted military advantage is indirect, long-term, or economic rather than concrete and direct. A facility’s contribution to national power does not automatically make every component a lawful military objective. The more attenuated the advantage, the heavier the burden on the attacker’s legal analysis.

Schools and health facilities collapse the distance between target categories and civilian life

The reported school strike is legally different from a grid node not because schools can never lose protection, but because the baseline protection is obvious and the evidentiary burden for military use is correspondingly grave. If a school is being used for military purposes, that fact matters. If it is not, attacking it is not an attack on infrastructure with incidental civilian consequences; it is an attack on a civilian object.

Health facilities require similar care. The Iranian Red Crescent site-damage figures reported 236 health facilities among civilian sites struck, but the figure alone does not answer whether each incident involved direct attack, incidental damage, military use, misidentification, or secondary effects.[5] It does identify the category of harm that IHL treats with particular seriousness. When medical services are degraded, the legal relevance is not limited to the day of the strike. Patients who cannot be moved, civilians who cannot obtain emergency care, and staff who cannot safely repair or operate facilities become part of the foreseeable harm picture.

That is why the ICRC’s focus on reverberating effects is not humanitarian atmosphere around the law; it is part of the legal work. In an urbanized, networked society, damage to electricity, water, and health systems can produce civilian harm through dependency chains rather than only through fragmentation and blast. A proportionality analysis that ignores those chains is incomplete.[7]

Public threats can matter, but they do not prove the targeting file

Public rhetoric is not a substitute for the targeting record. A lawful strike can be described recklessly by a politician, and an unlawful strike can be wrapped in careful language. Still, public statements can matter when investigators later ask about intent, knowledge, policy, and command climate.

The reported statements attributed to President Trump, including “stone ages” and “just for fun” comments, and Secretary Hegseth’s reported “no quarter” directive, are legally relevant for that limited reason.[3] They may help frame questions about whether civilian harm was treated as an unfortunate incidental risk, an intended pressure mechanism, or an acceptable method of coercion. They do not, by themselves, establish the elements of a war crime.

Luis Moreno Ocampo, the ICC’s founding prosecutor, drew attention to the accountability significance of public threats in an interview with The Harvard Crimson, connecting Trump’s threats to the logic behind ICC scrutiny of Russian officials Sergei Shoigu and Valery Gerasimov.[13] The analogy should not be overstated. Ukraine-related ICC proceedings arise in a different jurisdictional setting. But the underlying evidentiary lesson is sound: public threats against civilian infrastructure can become part of the file if attacks later appear to implement them.

A competent investigation would still need the materials the public does not have: target nomination records, collateral damage estimates, legal reviews, intelligence summaries, weaponeering choices, after-action assessments, battle damage assessments, civilian casualty reviews, and communications among commanders and civilian leaders. Without those records, the public legal assessment remains provisional. With them, some incidents might look more defensible than initial reporting suggests, and others might look worse.

The Rome Statute theory is serious; the jurisdictional path is narrow

If a court had jurisdiction, the Rome Statute would make several theories relevant. Article 8(2)(b) covers, among other things, intentionally directing attacks against civilian objects, intentionally launching disproportionate attacks, and certain attacks affecting protected objects and humanitarian interests. The reported strikes on schools, health facilities, water systems, and energy infrastructure would not all fit the same theory. Some would turn on whether the object was civilian. Some would turn on proportionality. Some would turn on whether the attacker took feasible precautions. Some would turn on intent.

The hard wall is jurisdiction. The United States is not a party to the Rome Statute and rejects ICC jurisdiction. The research record also indicates that neither the United States nor Iran has accepted ICC jurisdiction for this conflict. That leaves the ordinary ICC route largely unavailable unless jurisdiction can be grounded through some other accepted basis, which is not apparent from the public materials.

A UN Security Council referral would be the cleanest way to overcome that gap. It is also the least realistic in the near term. Any referral involving alleged US conduct would face the political reality of the US veto. That does not make the legal allegations unserious. It means that the international criminal architecture is structurally weakest when allegations concern a permanent member or its close strategic interests.

Universal jurisdiction is a narrower possibility. Some national systems allow war-crimes investigations even when the alleged crimes occurred abroad and the accused are foreign nationals. In practice, those cases face evidentiary barriers, immunities, political pressure, custody problems, and prosecutorial discretion. They are more plausible for lower-level perpetrators who travel into a cooperative jurisdiction than for senior US officials or commanders shielded by power, position, and domestic politics.

Domestic US law is not irrelevant. The US War Crimes Act and military justice mechanisms can, in theory, address grave breaches and certain war crimes. The difficulty is institutional will. Domestic accountability for a major national security campaign would require investigative independence, access to classified materials, and a willingness to examine senior decision-making rather than only isolated battlefield error.

The International Court of Justice presents a different path because it concerns state responsibility rather than individual criminal liability. But ICJ jurisdiction depends on consent, compromissory clauses, or other jurisdictional hooks. Even where a claim can be framed, remedies are slower, more diplomatic, and less directly connected to individual punishment. The ICJ can clarify law and responsibility; it cannot replace a criminal forum.

What would make the legal assessment stronger or weaker

The public record already supports serious concern. It does not support treating every reported strike as adjudicated. The difference matters. Legal professionals should resist both habits: the habit of sanitizing essential-services attacks with operational vocabulary, and the habit of announcing criminal liability before the evidentiary record can bear it.

• Illegality arguments would be stronger with evidence of repeated strikes on civilian-dependent systems supported by weak or generalized military-advantage claims.
• They would also be stronger with proof that officials knew foreseeable civilian outages, water failures, or medical collapse would be severe and prolonged.
• They would be stronger still if feasible alternatives were rejected despite reducing civilian harm without sacrificing comparable concrete military advantage.
• Illegality arguments would be weaker with verified military use of specific sites, reliable intelligence, careful proportionality assessments, and meaningful precautions.
• They would also be weaker with prompt civilian-harm reviews, repair facilitation, humanitarian access, and operational changes after unexpected reverberating effects.

The repair question deserves more attention than it usually receives. After a strike, the legal and humanitarian consequences do not stop at the crater. Engineers, utility workers, medical staff, civil defense teams, and local administrators inherit the operational reality. If they cannot reach the site safely, obtain spare parts, restore power, restart pumps, or move patients, the civilian harm continues. Those consequences are not always fully foreseeable. Some are. The law expects commanders to care about that distinction before the weapon is released.

The institutional judgment

The strongest current legal concern sits at the intersection of three facts: the public jus ad bellum objection identified by a large group of international law scholars, the reported pattern of attacks affecting essential civilian systems, and public rhetoric that appears to loosen rather than reinforce the discipline required by IHL. None of those facts alone proves a war crime in a specific incident. Together, they justify a serious legal inquiry.

A defensible military lawyer can imagine lawful strikes on particular dual-use energy or communications assets. A defensible military lawyer should also be troubled by any theory that treats civilian survival systems as ordinary logistics nodes once a military connection has been asserted. The law of armed conflict was not designed to make modern infrastructure untouchable. It was designed to keep military necessity from swallowing civilian life.

On the current public record, the 2026 US-Iran critical infrastructure strikes present credible cases for violations of international law. The harder truth is that credible legal concern and practical accountability are now separated by jurisdictional walls: US non-membership in the Rome Statute, lack of accepted ICC jurisdiction, the likely absence of a Security Council referral, and the limits of domestic or universal-jurisdiction alternatives. In the present architecture, recognizing possible illegality is far easier than enforcing it.

References

1. 2026 Iran war, Britannica.
2. US war crimes in Iran: civilian infrastructure and international law school strike, The Guardian, April 3, 2026.
3. Is the U.S. threatening to commit war crimes in Iran?, NPR, April 1, 2026.
4. Airwars casualty data, Airwars.
5. Iranian Red Crescent civilian site damage data, Iranian Red Crescent.
6. Professors' Letter on International Law and the Iran War, Just Security.
7. When the lights go out: Protection of energy infrastructure in armed conflict, ICRC Humanitarian Law & Policy, April 20, 2023.
8. The Dangerous Rise of Dual-Use Objects in War, Yale Law Journal, 2025.
9. Ukraine Symposium - Attacking Power Infrastructure under International Humanitarian Law, Lieber Institute.
10. ICRC president: War on essential infrastructure is war on civilians, International Committee of the Red Cross.
11. US-Iran grid conflict, C4Defence.
12. The Law of War - Critical Infrastructure Cyber Threats, Baker Botts, March 2026.
13. Ocampo Forum Iran, The Harvard Crimson, April 7, 2026.