This blog post is part of a series of posts published by the Research In Public International Law Group.
Introduction
Recent US and Israeli military operations against Iran have prompted intense debates about the legality of such force under contemporary international law. US Secretary of State Marco Rubio claimed the United States acted to support Israel as the US President “knew that there was going to be an Israeli action. We knew that that would precipitate an attack against American forces”. Yet this combination of justifications does not support the actions taken.
The US justification rests on speculative assessments about what Iran might have done, rather than demonstrable evidence of an impending attack. This conclusion is echoed forcefully in a plethora of expert commentaries on various international law blogs (see e.g. Milanovic 2026, Adil Ahmad Haque 2026). Whilst Lieblich 2026 has provided an excellent examination of the double pre-emption and imminence discussion, one element that has, however, not been elaborated upon is that the attack by the US could be considered an ad hoc form of collective self-defence. This post argues that, absent a lawful intervention by Israel, the US would not be able to claim it is acting in collective self-defence.
This post starts by explaining the prohibition of the use of force and its exceptions, notably Article 51 of the United Nations Charter and the customary rule of anticipatory self-defence. It then warns against an expansive interpretation that includes pre-emptive self-defence before providing an analysis of a flawed potential collective self-defence claim.
Article 51 and the threshold requirement of an armed attack
The prohibition of the use of force is a cornerstone of the United Nations Charter (ICJ DRC v Uganda 2005, para 148), anchored in its Article 2(4). Force can only be used in two specific circumstances: when the Security Council authorises the use of force (Chapter VII), and in the exercise of the inherent right of individual or collective self-defence if an armed attack occurs (Article 51).
Some quarters would argue that, given that Israel had already been attacked by Iran (some time ago), the right of self-defence did not extinguish simply because time had passed (see Lovitky 2025 on Israel’s attack against Iran in June 2025 following Iran’s attacks in October 2024). The law is clear: a State must prove that resort to force is necessary to repel the attack. First, it is difficult to see how the current attack fends off an attack that is over a year old. Second, necessity is also measured with due regard to whether nonforcible means could be used successfully (see Adil Ahmad Haque 2026; Schmitt, Bridgeman and Goodman 2026). Israel did not even attempt negotiating with Iran; rather, it was the US that was in discussion with Iran and, according to Oman’s Foreign Minister, “peace [was] within […] reach”. Yet, the US ended such negotiations. A claim of necessity cannot thus be made.
Anticipatory self-defence: A narrow and disputed doctrine
Under customary international law, anticipatory self-defence is recognised as a lawful use of force. This narrow exception, stemming from the 1837 Caroline incident, allows force when the State can show “a necessity of self-defence, instant, overwhelming, and leaving no choice of means and no moment for deliberation”. The rule of thumb is that the State is simply using force “before it is too late to do so”, i.e., to enable it to mount an effective defence (see O’Meara 2022). As Schmitt 2025 puts it, this is a test of “last window of opportunity”. However, nothing suggests that Iran was on the brink of launching an attack against Israel or the US.
Modern interpretations of this rule have focused on the imminence of the attack, each State (e.g. the US, the UK, and Australia) adopting its own standard. The lack of a common understanding both among States and scholars reflects the controversial nature of the use of force in anticipatory self-defence. Among the indicators used to determine the lawfulness of such action (see O’Meara 2022) are the probability or likelihood of the attack, the nature of the threat, the immediacy of the threat, whether the use of force is the only reasonable choice in the given circumstances, whether the response is proportionate, etc. That said, an anticipatory self-defence claim rests on actual imminent threats, not hypothetical scenarios. In practice, much of this links back to the availability of accurate intelligence information (see Bachmann and Keinan 2025). Nothing in the few days before the Israel/US attacks suggests that Iran was more likely to attack Israel than at any other point. In fact, Iran was at the time in negotiations with the US (see interview with Oman’s Foreign Minister).
Rubio’s justification did not rely on evidence of Iranian mobilisation but on US predictions that Iran might retaliate for Israel’s own pre-planned actions. What is more, Rubio’s statement could be interpreted as inadvertently undermining the US claim by stating that the US knew that “there was going to be an Israeli action” (a claim later withdrawn). He did stress the Iranian military preparations as the catalyst for US fears. This is the opposite of imminence! This reinforces the claim that Israel lacked a basis to claim imminence.
Pre-emptive self-defence: An even narrower and more controversial doctrine
Since the Bush era, the concept of pre-emptive (sometimes also confusedly referred to as preventive) strikes has gained prominence, though it is highly controversial. Whilst the US seemingly acted under such a doctrine when it invaded Iraq in 2003, it argued that it did not accept such a theory (see discussion in Taft and Buchwald 2003); rather, in its view, the intervention was based on Iraq having violated a UN Security Council resolution (see Finucane 2025). The doctrine of preemptive self-defence was mainly developed in the context of the “war on terror” to enable the US to strike State and non-State actors in a “non-consenting state” (see Lederman 2016), i.e. a State unable or unwilling to put an end to that threat on its territory (see Schmitt 2025).
There is, nonetheless, a wider reach of this doctrine, notably with the purpose of such strikes being to destroy a State’s threatening capabilities. This is more often called preventive self-defence (see Israel’s position as explained in Friedman 2026). In 1981, the US joined other States in the UN Security Council (Resolution 487 (1981)) in condemning Israel’s attack on Osirak, an Iraqi nuclear facility and rejecting similar justifications (see Finucane 2025). It pointed out the “absence of any evidence that Iraq had launched or was planning to launch an attack that could justify Israel’s use of force. … [T]he presence in a State of the military capacity to injure or even to destroy another State cannot itself be considered a sufficient basis for the defensive use of force.” (Sofaer, US Legal Adviser, as cited in Lederman 2016). Replace Iraq with Iran in the above statement, and the irony of such a statement in light of the current events is not lost on the reader. This historical consistency severely undercuts Washington’s current position. Rubio indeed stated that the US intervened “to eliminate the threat of Iran’s shortrange ballistic missiles and the threat caused by their Navy particularly to naval assets”.
It is evident that such a (flawed) understanding of anticipatory self-defence differs significantly from the original interpretation of the Caroline incident formulation. For certain, the law evolves, but even the most contemporary interpretations of anticipatory self-defence do not authorise preventive strikes based merely on an adversary’s developing capabilities. As Schmitt 2025 emphasises, “States enjoy no right under international law […] to use force solely because another State is about to acquire a threatening capability”.
Collective self-defence
Under international law (both Article 51 UN Charter and customary international law (see ICJ Nicaragua 1986, para 194)), a State is allowed to seek support through collective self-defence, which denotes the situation whereby a State acts to defend an ally that has suffered an armed attack or is facing an imminent threat. Interestingly, the US support for Israel has not been couched in terms of collective self-defence. That being said, let’s entertain this claim here. For collective self-defence to be lawful, the self-defence must comply with the jus ad bellum rules (see Green 2024; Schmitt, Bridgeman and Goodman 2026). This means that if Israel lacked a valid claim of self-defence, then the United States cannot invoke collective self-defence.
First, it must be noted that the US did not refer to collective self-defence in this recent attack. This contrasts to the situation in June 2025, when the US had invoked it in its letter to the UN Security Council to justify its attack on Iran by first referring to Israel responding to Iranian attacks “in the exercise of its inherent right of self-defence” and adding that “[i]n close coordination with the Government of Israel, the United States has taken necessary and proportionate action […] to defend Israel and also to protect our own security, citizens, and interests, consistent with Article 51 of the Charter of the United Nations.” Even at the time, Finucane 2025 demonstrated that such a claim violated international law. As explained, Israel can certainly not claim to be acting in self-defence. This forecloses the possibility of a lawful “derivative” US collective self‑defence.
Second, from a procedural perspective, the victim State must declare that it has suffered from an armed attack, and it must request military aid in response (ICJ Nicaragua 1986, paras 165, 195 and 199). Israel therefore needs to issue a declaration that it has been attacked and then an ad hoc request for the US to join them in self-defence. Green 2017, however, points out that, in practice, the declaration requirement finds no relevance. In contrast, the request for assistance is (see Green 2024). No request was made by Israel; rather, it seems that the US intervened out of fear that it would be attacked (Rubio).
Third, and relatedly, the US rationale for its intervention appears to be also grounded in its own projection that Iran might retaliate against US assets following Israeli escalation (see Rubio). In other words, the putative attack was on the US and not Israel, which, following this logic, would make it impossible for the US to claim that it acted in collective self-defence to support Israel.
Conclusion
This interpretive drift regarding anticipatory self-defence has become particularly dangerous (see also Schmitt 2025). While governmental legal teams could be perceived to be bending doctrines to suit policy goals rather than grounding policy in legal constraints, it gives the impression that an increasing number of political leaders now see international law not only as a constraint but also as an utter nuisance. The US logic depends on a chain of hypothetical contingencies, reminiscent of the “Minority Report” scenario. International law categorically rejects such conjectural reasoning. In the end, the intervention reflects precisely what international law seeks to prevent: unilateral, pre-emptive uses of force dressed in the language of (collective) self-defence.