Iran’s Strikes on the Kurdistan Region: The Limits of Self-Defence Against Non-State Actors

[Arez Rizgari is an Iranian scholar. The author is writing under a pseudonym because they are concerned that this post will subject them to political retaliation or restrictions on their academic freedom by Iranian authorities.]

On the morning of 17 July 2026, the Islamic Revolutionary Guard Corps (IRGC) struck the offices of the Komala Party of Iranian Kurdistan in the Kurdistan Region of Iraq (KRI) with drones and missiles. At least ten Peshmerga were killed and a number of others wounded (here & here). According to the Kurdistan Human Rights Network, the missiles struck the reception area of the party’s headquarters at around 6:00 a.m. local time. The Kurdistan Regional Government (KRG) condemned the attacks and called on Iran to cease the escalation. The KRI presidency separately denounced attacks as a “blatant violation” of Iraq’s sovereignty.

The strikes persisted, including a drone attack on 19 July against one of the bases of the Kurdistan Freedom Party (PAK) that wounded at least nine more Peshmerga. These events cannot be viewed as isolated incidents. Since the conflict between Iran and the US-Israeli coalition erupted on 28 February 2026, Iran and its proxy militias have routinely targeted the KRI. According to data from the Community Peacemaker Teams, 751 such attacks took place by late May. Furthermore, the Hengaw Organization for Human Rights logged 167 targeted strikes against camps housing Iranian Kurdistan (Rojhilat) parties during the 40-day conflict. More casualties have been reported since the ceasefire on April 8.

The parties based in KRI, form part of the Kurdish liberation movement that has, for over a century, pursued the realisation of the Kurdish nation’s right to self-determination through both political and armed means. This struggle is rooted in the denial of the Kurds’ right to self-determination and the division of Kurdistan among five states (later four) after First World War, and the obstruction of the path to achieving this right within the state-centered international legal order. Setting aside this fundamental critique, the legal inquiry here is narrower: can Iran’s strikes on parties based in Iraq be justified under the jus ad bellum of existing state-centric international law?

Use of Force Without a Legal Justification

Iran has not made any official claim to self-defence. As far as the evidence shows, Iran has not sent any notification to the Security Council (SC) under Article 51 of the Charter regarding the 17 July strike, nor has it publicly invoked concepts such as self-defence, necessity, or other legal justifications for escalating attacks on these camps. Even the attribution of these strikes to Iran is based on a known pattern and the claims of the targeted parties, rather than on official admission by Tehran; as Shafaq News reports, Iran is not mentioned by name in the official count of these individual strikes by KRI security authorities, while the report itself considers the logic for these specific attacks to be “known” (Tehran settling scores with its exiled dissidents ). This silence, however, is not without precedent. 

In a letter dated 11 September 2018 to the SC, Iran’s representative acknowledged that Iran had carried out a “limited and proportionate military action” on September 8, 2018, specifically against the PDKI headquarters and leadership, citing that the party had been carrying out cross-border armed attacks on Iranian border areas for several years — including nine in just the previous four months — and that Iran had previously taken years of diplomatic actions before Iraqi authorities to stop these activities. This letter makes two points for the present analysis: First, Iran knows exactly how to substantiate a claim of self-defence by invoking Article 51, describing the specific threat, and referring to previous diplomatic action — a practice that has not been repeated for the current case. Second, even Iran itself, characterised such action as “limited and proportionate” and limited to the “headquarters and leadership” of a specific target — a description that does not fit well with a months-long campaign against dozens of fixed targets across the region.

States, even when they violate international law, generally attempt to justify their conduct. But a state that makes no attempt to justify its use of force on another state’s territory has effectively shown that it has not troubled itself to bring its conduct within the Article 51 exception — and absent that exception, this remains a violation of Article 2(4) without any legal defence. This silence is therefore a decisive legal proposition in itself. Legal commentators seeking to reconstruct the strongest possible legal justification for this state’s actions should not deliberately fill this gap with their own assumptions.

The Requirements of Self-Defence

The traditional interpretation of Article 51 — and the one endorsed in the jurisprudence of the ICJ (notably Nicaragua, para. 195, and Wall, para. 139) — remains wedded to a state-centric paradigm, confining self-defence to attacks attributable to a state. Moreover, any customary basis for anticipatory self-defence, at best, remains confined to the Caroline formula’s threshold of imminent and overwhelming threat. Expansive interpretations, by contrast, extend both the temporal dimension of the “attack” (by admitting preemptive self-defence) and the identity of the attacker (by admitting the possibility of an attack by non-state actors). Despite some state support, these expansive readings remain contested both in their temporal dimension and with respect to the identity of the attacker.

The public record does not establish that Iran faced an imminent and overwhelming attack by Kurdish parties on its territory. For completeness, and using only what is already in the public record, within the framework of a broad interpretation, it might be said that the best explanation — perhaps even justification — for Iran’s operations lies in reporting on the February–April war indicating discussions within the Trump administration about arming Rojhilat parties for entry into Iranian territory. Two considerations, however, preclude reading this as a justification for the 17 July strike. First, regardless of the impact of this event on strikes carried out during the 40-day war, the 17 July strike occurred more than three months after the ceasefire; while the claim of self-defence must be justified by the threat existing at the time of the attack, not by the situation three months earlier. Second, this plan itself never reaches the threshold required for an actual or imminent armed attack. Reports indicate the plan was twice readied for launch before being called off amid leaks and allied pressure, yet no coordinated attack or transfer of weapons was ever confirmed — falling short of an actual or imminent armed attack.

It might further be argued that the pattern of these parties’ own conduct over time — irrespective of any external support — meets the threshold of an armed attack under the “accumulation of events” doctrine (Nicaragua, para. 231 and Oil Platforms, para. 64). This doctrine refers to a series of successive attacks connected in time, origin, and cause, particularly where they form part of an “overall and coordinated campaign” (Ruys, p. 168), and Iran has previously invoked it in relation to strikes on Mojahedin-e-Khalq camps in Iraq (e.g. here and here). But no comparable public evidence documents any such pattern of successive, connected attacks by these parties against Iran, as part of an overall campaign; this argument, too, fails on the facts of the present case.

Moreover, the actions taken fall well short of the requirements of necessity and proportionality (Nicaragua, para. 194). Self-defence against a non-state actor, on any account, is a response to a specific or imminent armed attack — not a general licence to dismantle an opposition movement’s political and organisational infrastructure over time. No evidence in the public record connects the 17 July strike on Komala’s headquarters to any specific, ongoing, or imminent armed attack from that camp. What the record shows instead is a broad and repeated pattern by Iran over time: 751 attacks on the Region in three months; 167 strikes during the 40-day war alone, with further deaths after a ceasefire that appeared to have restored calm. An operation of this duration and scale, directed at fixed party headquarters rather than forces engaged in active hostilities, looks less like a series of necessity-driven defensive responses and more like an open-ended effort to suppress these parties as such — something Article 51, on any reading, does not authorise.

The “Unwilling or Unable” Doctrine

The “unwilling or unable” doctrine has, in recent decades, been a frequently cited basis for strikes against non-state actors operating in the territory of a third state. It holds that a state may use force against non-state actors in a third state’s territory where that state is unwilling or unable to suppress the threat itself. This doctrine, however, has not yet become established as a customary rule and remains contested; some authors reject its legitimacy altogether, while even its proponents maintain that it currently lacks sufficient content to serve as a restrictive international norm.

Even taking this doctrine as the governing framework, the real picture is more complicated than it first appears. To give this standard workable content, Deeks proposes six factors that a victim state should weigh before resorting to force (pp. 519–532); according to the facts of the present case, only one factor clearly favours Iran. 

Iran has failed to seek Iraq’s consent for its use of force on Iraqi territory — the first factor Deeks identifies (p. 519). Nor does anything in the public record establish the scale, sophistication, or imminence of further attacks by these parties that would, on Deeks’ own account of the second factor (p. 521), make a demanding assessment of the KRG’s capacity understandable. Iran has, in recent years, repeatedly warned the KRG of the threat of missile strikes against these parties; it publicly demanded, from 2021, that Iraq expel them; in March 2023 it signed a formal security agreement with Iraq’s central government committing Baghdad to disarming and relocating these parties; and Nechirvan Barzani, on a subsequent visit to Tehran, pledged that Erbil would work to disarm these parties and expel them from the KRI. This arguably satisfies the third factor of the Deeks framework — a request to address the threat and adequate time to respond (p. 521).

According to a Soufan Center report, the KRG — unlike a state that tolerates or facilitates attacks on a neighbour — has since directly communicated and urged them not to carry out cross-border operations into Iran from KRG territory, and has arranged for the closure of these parties’ camps and the handover of their heavy weapons to the KRG. What remains, years after that agreement, is neither Iraq’s outright refusal nor its full compliance, but a formal commitment with incomplete implementation — precisely the situation for which Deeks’ fourth factor, a “reasonable assessment of the territorial state’s control and capacity” (p. 525), is designed. Notably, Iran itself implicitly conceded this shortfall at the time: when Iraq’s foreign minister claimed full implementation of the agreement in October 2023, Mohammad Hossein Bagheri, Chief of Staff of Iran’s Armed Forces, responded that all that had occurred in those six months was that these groups had “just distanced a bit from the borders of our country”. Whether, three years on, the KRG genuinely lacks the capacity to fully implement this commitment is an empirical question Iran has never publicly and evidentially answered on an up-to-date basis. The mere existence of a signed agreement and the repetition of demands does not, by itself, establish that the KRG remains “unable.” Discharging that burden through a reasoned assessment falls to Iran — a burden it has not attempted to meet. Furthermore, Iran has failed to show why the KRG’s proposal — to close the border camps and disarm these parties — fell short of what Deeks’ fifth factor expects from “a reasonable state” in a similar position (p. 529); on the public record, this looks closer to a plan that a reasonable state would accept than one it would dismiss as illusory. Ultimately, the KRG’s record of closures, disarmament, and relocation directly undermines, rather than supports, any inference of passivity under the sixth factor (p. 531).

These events also raise points from the perspective of jus in bello. Komala’s statement on the strike described all of those killed as “Peshmerga recruits who were preparing to join the ranks” of its forces, not yet inducted members — consistent with Hengaw reporting that several were volunteers undergoing the party’s admission process, and HANA reporting several had only recently joined with no verified evidence of a “continuous combat function.” Under the ICRC’s Interpretive Guidance, an individual’s status turns on whether they have assumed a continuous function of direct participation in hostilities — a status that presence in an administrative reception processing, or a party’s own honorific description of its fallen recruits, does not establish; it is this criterion, not the label “Peshmerga” or “party member” alone that determines whether an individual is immune from direct attack. Hengaw has, further, reported white phosphorus use in the drone strike on the PAK bases. Should these findings be confirmed, they raise separate legal questions concerning compliance with the principles of distinction, proportionality, and precaution in these specific attacks — questions that lie beyond, and independent of, the analysis in this post.

Conclusion

None of the legal arguments examined above resolve the previous and separate question of the right to self-determination of the Kurdish nation — a question that this piece has deliberately not addressed. But that distinction is precisely the point; within the constraints of this state-centric framework, the legality of Iran’s strikes on Iraqi territory does not depend on the merits of Kurdish claims, and should not be allowed to turn on them. Under jus ad bellum, Iran’s military operations against Kurdish party camps in northern Iraq rest on no international legal justification and amount to a violation of Article 2(4). A state that strikes another state’s territory roughly 750 times without providing a legal explanation for its actions, is not seeking to test the limits of self-defence — it is simply operating outside it; in a zone where the absence of consequence and international response risks converting violations into precedent.