Why the Putin ICC Arrest Warrant Has Failed Three Times

By the time Tajikistan hosted Vladimir Putin in October 2025 without arresting him, the problem had stopped looking like an awkward first test of a difficult warrant. It had become a sequence. South Africa had sought consultations before a 2023 summit and avoided the confrontation when Putin stayed away. Mongolia then hosted him in September 2024 and received the first formal non-compliance finding. Tajikistan repeated the hosting pattern, invoked regional immunity commitments, and was called out by the ICC in March 2026 for failing to arrest him despite the warrant.[1][2]

That sequence matters more than any single diplomatic episode. The legal question has never been only whether the International Criminal Court can issue a warrant for a sitting president of a non-party state. The sharper question is what the Rome Statute system does after a state party declines to execute it. As of Q3 2026, the answer is thin: a finding, a referral, a Bureau process, and no meaningful penalty visible enough to change the risk calculation for the next government weighing arrest against bilateral cost.

The Three Episodes Are Not the Same, but They Point in the Same Direction

South Africa is the cleanest example of near-non-compliance rather than completed non-compliance. Before the August 2023 BRICS summit, it entered Article 97 consultations with the ICC, the mechanism used when a requested state sees problems that may impede execution of a cooperation request. The ICC’s confidential response, as later discussed in legal analysis of the episode, confirmed South Africa’s obligation to arrest Putin if he entered its territory. Putin ultimately did not attend the summit in person.[3]

That detail is important because South Africa did not merely face a vague political demand from The Hague. It asked the Court about the legal impediment and received the answer a practitioner would care about: the arrest obligation remained. The arrest scenario was then avoided by itinerary, not resolved by enforcement.

Mongolia removed the ambiguity. It is a Rome Statute state party. Putin traveled there in September 2024. Mongolia did not arrest him. In October 2024, ICC Pre-Trial Chamber II found that Mongolia had failed to cooperate in the arrest and surrender of Putin and referred the matter to the Assembly of States Parties under Article 87(7).[4]

The Mongolia finding is the most legally useful of the three episodes because it forced the Court to restate the immunity point in the posture that matters: not an academic exchange about sitting heads of state, but a state-party failure after the wanted person had been physically present. The Chamber reaffirmed that head-of-state immunity was not opposable before the ICC and that no waiver from a third state was required for the Court to proceed with its request.[4]

Tajikistan then made the issue harder to treat as a singular diplomatic embarrassment. It hosted Putin in October 2025 and did not arrest him. The argument reported in the non-compliance dispute was not simply a reprise of ordinary head-of-state immunity; Tajikistan invoked Commonwealth of Independent States commitments as an immunity-based objection, a variant of the Article 98 problem that state parties have raised when they say other international obligations prevent execution of an ICC request.[1][2]

The ICC’s March 2026 response again moved the matter into the non-compliance channel, escalating it to the Assembly of States Parties Bureau.[1][2] That escalation should not be mistaken for enforcement. It is an institutional handoff. The receiving institution is political, consensus-heavy, and, on the materials available, not equipped with a sanctions toolkit that would make non-compliance expensive in the ordinary sense of the word.

The Warrant’s Legal Foundation Has Not Collapsed

Calling these episodes failures does not mean the Putin warrant is legally unserious. The ICC Pre-Trial Chamber II warrant was issued on March 17, 2023. It concerns alleged responsibility for the unlawful deportation and transfer of children from occupied areas of Ukraine, with the Rome Statute provisions identified as Article 8(2)(a)(vii) and Article 8(2)(b)(viii), and modes of liability under Articles 25(3)(a) and 28(b).[4]

The jurisdictional basis also has to be kept straight. Russia is not a Rome Statute party. The Court’s Ukraine situation rests on Ukraine’s Article 12(3) ad hoc declarations, made in 2014 and 2015, accepting ICC jurisdiction over relevant alleged crimes committed on Ukrainian territory.[4] Whether one approves of the Court’s broader institutional role is not the issue. The warrant sits inside an asserted jurisdictional framework, and state parties receiving cooperation requests are not free to treat the request as merely expressive.

The immunity question is where legal and operational audiences often talk past each other. Within the ICC’s own doctrine, the leading authority remains the 2019 Al-Bashir Appeals Chamber approach, under which head-of-state immunity does not bar surrender to the Court in the relevant circumstances. Scholars have criticized aspects of that reasoning, including the way it relied on prior international tribunal jurisprudence rather than a systematic survey of state practice and opinio juris. But that critique does not give a state party a practical permission slip to ignore an outstanding ICC cooperation request. It means the doctrine is contested in academic and intergovernmental terms, not that the warrant disappears at the border.

For legal-risk purposes, that distinction is not decorative. A valid warrant can still be a weak enforcement instrument. A contested immunity theory can still be the Court’s controlling position. A state can be found non-compliant and still suffer no immediate material consequence. Precision is not a luxury here; it is the difference between advising on law and advising on wishful thinking.

Article 87(7) Sends the Problem Somewhere Else

Article 87(7) is often described as if it were the Court’s enforcement lever. It is more accurately a reporting mechanism. When a state party fails to comply with a cooperation request, the Court may make a finding of non-compliance and refer the matter to the Assembly of States Parties. That is what happened with Mongolia, and that is the channel into which Tajikistan was later moved.[4][1]

The legal form has value. It preserves the record, prevents quiet normalization, and tells other states that the Court has not accepted the immunity objection. But a finding is not a fine, a suspension, a trade measure, a travel ban, or a loss of voting rights. It is not self-executing. It does not put marshals on an airport tarmac. It is a legal conclusion handed to a political body.

That handoff is where the Putin warrant exposes the system’s least flattering feature. The Assembly of States Parties has no real sanctioning power against non-compliant states, and commentary on the enforcement crisis has noted that Article 87(7) referrals have not produced tangible consequences against non-compliant states.[5] The point is not that every failure should trigger a theatrical punishment. The point is that repeated non-compliance without visible cost teaches future requested states what the procedure is worth.

A compliance system can survive occasional breach. It struggles when breach becomes administratively legible and still consequence-light. South Africa demonstrated that a state could seek clarification and maneuver away from the arrest. Mongolia demonstrated that a state could host the target and absorb a formal finding. Tajikistan demonstrated that a second state could do the same after Mongolia had already been referred. The direction of travel is not subtle.

The Mongolia Finding Did More Legally Than Politically

Mongolia deserves the most attention because it produced the cleanest record. The Court did not need to speculate about a hypothetical future visit. Putin entered a state party’s territory. The arrest request was not executed. The Chamber then issued a formal finding and referred the matter to the ASP.[4]

As law, that matters. It avoids the worst institutional outcome, which would have been silence after an open breach. It also locks in the Court’s position that a sitting head of state from a non-party state cannot rely on personal immunity to defeat the Court’s request when a state party is asked to cooperate.[4] Future state parties may dislike that conclusion, but they cannot plausibly say they were not warned.

As enforcement, however, Mongolia shows the limit of restatement. The Court spoke; the requested arrest had already failed. The referral placed the matter before the Assembly of States Parties; the record supplied here does not identify any meaningful penalty imposed on Mongolia. If the next state party asks what happened after Mongolia declined to arrest Putin, the operational answer is not encouraging.

This is why the common symbolic-accountability framing is incomplete. A warrant can narrow diplomatic space, complicate travel planning, and attach legal stigma. Those are not trivial effects. But where a state party actually receives the wanted person and does not arrest him, the system has to be judged by what follows. If the follow-on consequence is only a file moving from a Chamber to the ASP, the deterrent signal is modest.

Tajikistan Added a New Immunity Route, Not a New Enforcement Answer

Tajikistan’s case is not just Mongolia with a different flag. The reported objection based on CIS immunities matters because it shows how non-compliance arguments can migrate from classic personal immunity into treaty-conflict terrain. Article 98 of the Rome Statute is the provision states tend to invoke when they argue that compliance with an ICC request would require them to act inconsistently with certain international obligations regarding immunity or consent.[1][2]

That does not mean Tajikistan’s argument succeeds. It means the Court is facing not only defiance but legal packaging. For the next state party considering whether to host a wanted official, the playbook is no longer limited to saying that sitting presidents are immune. It can include regional agreements, bilateral obligations, or other asserted conflicts that make non-execution appear lawyered rather than merely political.

The March 2026 escalation to the ASP Bureau therefore matters as a warning sign. It shows that the Court will continue to press the obligation. It also shows, again, that the Court’s immediate remedy is procedural escalation rather than coercive enforcement.[1][2] For legal professionals, the distinction is not pedantry. It affects how one reads the warrant as a risk signal: serious as law, uncertain as compulsion.

The Assembly Gap Is Now Part of the Risk Assessment

The Rome Statute system was not built with an international police force. It depends on state cooperation. That dependency is not a hidden footnote; it is the architecture. But architecture becomes a governance problem when the same failure mode appears repeatedly and the political body receiving referrals does not impose visible costs.

The St Andrews Law Review assessment puts the enforcement problem in direct terms: the ASP lacks sanctioning power against non-compliant states, and the Article 87(7) referral mechanism has produced no tangible consequence against any non-compliant state.[5] That is the compliance fact that should sit beside the warrant itself in any institutional-risk memorandum.

There is a tendency, especially in public commentary, to treat the ASP referral as if it were the punishment. It is not. It is the point at which the Court has exhausted its immediate judicial response and asks the community of states parties to act. If that community responds with concern, dialogue, or silence, the underlying arrest obligation remains formally intact but operationally weakened.

The consequence gap also changes how future requests are received. A state party weighing arrest may still fear domestic litigation, reputational criticism, diplomatic friction with other states parties, or damage to its standing with the Court. Those pressures are real but uneven. They do not substitute for an institutional penalty that attaches predictably to non-compliance.

For a powerful-state target, that unevenness is the point. The harder the arrest is politically, the more the system needs dependable follow-through. If the follow-through is discretionary and consequence-light, then state parties can convert the warrant into a matter of diplomatic tolerance: worth honoring when convenient, worth evading when costly.

The Peace-Conference Decision Should Be Read Narrowly

A further complication arrived in June 2026, when the ICC reportedly opened a narrow pathway for a wanted head of state to attend a UN-sponsored peace conference without facing immediate arrest. The best public reading currently available comes from Gleb Bogush’s Moscow Times analysis, but the relevant ICC record, identified as ICC-01/22-146-Red, remains partially classified and the requesting state has not been publicly identified.[6]

That partial record matters. The decision should not be inflated into a general waiver, a suspension of the Putin warrant, or an abandonment of the Court’s arrest position. It is better understood, on the currently available information, as a narrow peace-process accommodation whose full reasoning cannot yet be checked against the complete record.[6]

Still, narrow exceptions can affect compliance expectations. If states already see that non-compliance referrals carry little visible cost, and then see even a limited pathway for conference attendance, they may read the system less as a hard arrest regime than as a set of obligations mediated by context. Lawyers should resist overstating that point. They should also not ignore it.

What the Warrant Now Signals

For a legal analysis of the Vladimir Putin ICC arrest warrant, the clean answer is unsatisfying but necessary: the warrant has not failed as law; it has failed three times as an arrest mechanism. South Africa did not execute it because the visit did not happen. Mongolia hosted Putin and did not arrest him. Tajikistan hosted him and did not arrest him. The ICC responded within its procedural vocabulary. The ASP process has not supplied the missing coercive consequence.

That leaves legal professionals with a bifurcated risk signal. On one side, the warrant is legally serious. It affects travel planning, diplomatic optics, institutional relationships, and any state-party interaction where arrest could plausibly be demanded. It also tells courts, ministries, and counterparties that the ICC has made a formal criminal-law claim against the sitting Russian president.

On the other side, the enforcement record is poor where arrest would require a state party to absorb serious political cost. The system relies on voluntary cooperation and then, after refusal, on a political assembly without meaningful sanctioning power. That is not the death of international criminal law. It is a concrete institutional limit, and it is now evidenced by more than one episode.

The prudent conclusion as of Q3 2026 is therefore neither triumphalist nor dismissive. The Putin warrant remains a valid and grave instrument within the ICC’s legal framework, but its practical force depends on state-party willingness to choose arrest over diplomacy, security ties, energy relationships, regional commitments, and political convenience. Repeated non-compliance without meaningful consequences has become part of the warrant’s legal environment, not an external embarrassment to be bracketed away.

References

1. ICC calls out Tajikistan over failure to arrest Putin despite warrant, Courthouse News Service.
2. Tajikistan’s failure to arrest Vladimir Putin undermines the fight against impunity, Coalition for the International Criminal Court.
3. The Putin South Africa Arrest Warrant Saga: A Tale of the Shrinking World of an Accused War Criminal, EJIL:Talk.
4. Ukraine situation: ICC Pre-Trial Chamber II finds Mongolia failed to cooperate in the arrest and surrender of Vladimir Vladimirovich Putin and refers the matter to the Assembly of States Parties, International Criminal Court.
5. The Enforcement Crisis of International Criminal Law: The Case of the Putin Arrest Warrant, St Andrews Law Review.
6. Did the International Criminal Court Just Ease Restrictions on Putin?, The Moscow Times, June 23, 2026.