The machinery of war crimes accountability has rarely looked busier, or more openly defied, than it did from 2025 into May 2026. International judges convicted a former militia commander for atrocities in Darfur[s], a former head of state was surrendered to The Hague to face crimes-against-humanity charges[s], and national courts in twenty countries logged dozens of fresh prosecutions[s]. Then, in mid-December, a Moscow court sentenced the International Criminal Court’s own prosecutor to fifteen years in prison[s], and the United States sanctioned two judges who had voted to reject Israel’s challenge to the Gaza investigation[s]. Earlier, Hungary began a withdrawal from the treaty that created the Court after hosting Benjamin Netanyahu, only for its government to discontinue that process in May 2026[s].
That contradiction is the central fact of international justice today. The courts can investigate, indict, try and convict. What they cannot do is make an arrest. Every stage of war crimes accountability, from the knock on the door to the cell where a sentence is served, runs through the police, prosecutors and prisons of sovereign states, and that chain holds only as long as a government chooses to honour it. When a state refuses, there is no bailiff to send in its place.
How the machinery of accountability is built
Modern international justice grew out of catastrophe. After the Yugoslav wars and the Rwandan genocide, governments created ad hoc tribunals, the ICTY and the ICTR, as one-off responses to specific atrocities[s]. Those courts advanced the law and convicted senior perpetrators, but a European Parliament study notes their work was repeatedly undermined “by high costs, lengthy proceedings, selective enforcement and lack of state cooperation”[s]. The lesson drove states to build something permanent: the International Criminal Court, established in 2002 under the Rome Statute[s].
The Court was never meant to handle every case. Under the principle of complementarity, “States have primary responsibility for the prosecution of war crimes,” and the ICC “only steps in when a State appears to be either unwilling or unable to prosecute”[s]. Alongside it runs a second tier: universal jurisdiction, which “enables national courts to pursue perpetrators regardless of nationality or location” and has revived in Europe through trials tied to Syria and Ukraine[s]. The result is less a single world court than a layered approach to war crimes accountability: national courts, hybrid panels and the ICC, each meant to catch what the others miss.
None of it is free, and the money exposes the politics. The European Union’s member states together are “the ICC’s largest financial contributor, providing roughly half of its annual budget”[s]. A court that leans on a handful of friendly governments for half its funding is a court whose survival depends on their continued goodwill, which is one reason the politics of war crimes accountability are written plainly into its budget.
Why war crimes accountability begins and ends with an arrest
For all its layered design, the system has one chokepoint. As the International Bar Association puts it, “without arrest and surrender, there can be no ICC trials. And without trials, there can be no justice or reparations for victims”[s]. The Court issues a warrant, but it has no marshals to execute it. “Without its own law enforcement, the ICC relies wholly on the national authorities of states,” above all the 125 governments that have joined the Rome Statute, “to arrest suspects and surrender them”[s].
The Court’s own record shows how often that hand-off fails. Since 2002 it has publicly issued 65 arrest warrants and executed only 24 of them; as of 1 October 2025, 32 suspects remained at large[s]. Roughly 63 percent of those public warrants, 41 of 65, have never been carried out. A warrant that is never enforced is, in practical terms, a press release with a seal on it.
This is also where the law collides with head-of-state immunity. Article 27 of the Rome Statute “explicitly rejects official capacity as a bar to jurisdiction,” yet in practice that principle “frequently clashes with state practice and political hesitation surrounding customary international law regarding personal immunity”[s]. On paper a sitting president enjoys no protection from the Court. In a foreign capital, arresting one is another matter entirely.
When states simply say no
2025 supplied a casebook of refusals. In January, Italian police arrested a Libyan general, Osema Almasri Najeem, who was wanted on an ICC warrant for alleged war crimes and crimes against humanity tied to a Tripoli prison. Within two days he was released on procedural grounds and, in the words of one legal review, “immediately expelled by a decree of the Minister of the Interior and repatriated to Libya onboard a State flight”[s]. The Court opened non-compliance proceedings against Italy under Article 87(7) of the Statute[s]. Italy’s foreign minister, Antonio Tajani, was blunt about why a bigger target would never be cuffed, asking of the warrant against Israeli Prime Minister Benjamin Netanyahu: “how could we arrest him? Can you imagine? Shall we engage in a firefight with the Mossad?”[s]
That warrant was not hypothetical. In November 2024 the Court issued arrest warrants for Netanyahu and his former defence minister, Yoav Gallant, for alleged war crimes and crimes against humanity in Gaza[s]. When Netanyahu visited Budapest in April 2025, Prime Minister Viktor Orbán’s government not only declined to arrest him but announced it would leave the Court altogether. Hungary began the Article 127 withdrawal process, and the withdrawal was set to take effect on 2 June 2026; one January 2026 analysis treated that pending exit as proof that “States Parties can exit from inconvenient obligations without consequence”[s][s]. But that last point did not hold: on 27 May 2026, Hungary’s parliament voted to remain in the ICC, and the Assembly of States Parties welcomed the decision to discontinue the withdrawal process[s][s]. Before Hungary reversed course, the Assembly of States Parties had responded by holding, for “the first time in 24 years,” a dedicated session on non-cooperation[s]: the first formal admission, after more than two decades, that defiance had become systemic. Each refusal narrows war crimes accountability to the cases powerful governments are willing to tolerate. These are not isolated lapses, but symptoms of a deeper problem in international law enforcement, which has courts and treaties in abundance and almost no way to compel a powerful state to comply.
When the enforcers become targets
By late 2025, major powers were no longer merely ignoring the Court but punishing the people who run it. On 12 December, a Moscow court issued in absentia verdicts against ICC Prosecutor Karim Khan and eight judges, sentencing Khan to fifteen years and the judges to terms ranging from 3.5 to 15 years for their role in warrants tied to Putin[s]. Weeks later Vladimir Putin “signed legislation blocking the enforcement of rulings by foreign criminal courts in Russia”[s], formalising a refusal that began when the Court issued 2023 warrants for him and his commissioner for children’s rights, Maria Lvova-Belova, over the alleged deportation of Ukrainian children[s].
The United States reached for financial weapons instead. After appeals judges rejected Israel’s challenge to the Gaza investigation, Washington sanctioned the two who voted with the majority[s]; an executive order earlier in the year had already authorised “asset freezes and entry bans on ICC officials”[s]. The effect is personal and near-total. One sanctioned French judge, Nicolas Guillou, “reported immediate loss of access to Amazon, Airbnb, PayPal, Visa, and MasterCard,” a punishment the legal scholar Kai Ambos called “a form of civil death”[s]. Human Rights Watch’s international justice director, Liz Evenson, framed the campaign plainly: “Government efforts to undermine the ICC reflect broader attacks on the global rule of law, aiming to disable institutions that seek to hold those responsible for the worst crimes to account”[s]. The campaign made war crimes accountability personally costly for the very people meant to deliver it.
None of it can be undone in the one chamber built to enforce the Court’s referrals. Where a permanent member is the target or the patron, a Security Council veto turns collective enforcement into a dead letter, and the referrals on Sudan and Libya have languished for exactly that reason.
Where war crimes accountability still works
The grim arithmetic at the top of the system obscures a quieter success below it. While the ICC’s marquee warrants gather dust, national courts have become the workhorses of war crimes accountability. A coalition of human-rights groups counted “91 extraterritorial and universal jurisdiction cases prosecuted in 20 countries” in 2025, with “34 new cases were opened or made public, and 23 convictions were secured”[s]. Four states, Kosovo, Peru, Poland and Türkiye, exercised that kind of jurisdiction for the first time[s].
These are mostly cases the headlines miss: lower-ranking perpetrators tried far from where they offended, often under the grave-breaches rules of the Geneva Conventions that oblige a state to prosecute or extradite. The work is open-ended. An investigator from the Dutch international crimes unit, whose Hague courtrooms hear many of these trials, observed that “even if all armed conflicts worldwide were to cease today, the unit would still face decades of work arising from investigations that are already underway”[s]. This is where most real punishment now happens, one defendant at a time, in domestic dockets rather than international ones.
Can the machinery be repaired
Reformers are not short of ideas, only of leverage. The International Bar Association’s November 2025 report sets out thirty recommendations for closing the arrest gap, from stronger national surrender laws to a dedicated capacity for tracking suspects[s]. The European Union holds tools of its own, including a blocking statute it could activate to blunt the US sanctions, but it “has yet to act to use its blocking statute”[s].
Other proposals tackle the politics head-on. Article 16 of the Rome Statute already lets the UN Security Council “defer ICC investigations or prosecutions for renewable periods of twelve months”[s], a pressure valve meant to stop arrest warrants from derailing peace talks, though it has rarely been used as designed. The deeper claim is that the choice between courts and negotiations is a false one. As Kofi Annan argued, in words a professor of international law recently put back before a Strasbourg audience, “justice and peace are not contradictory forces. Rather, properly pursued, they promote and sustain one another”[s]. Whether war crimes accountability survives the decade will depend less on the law than on the will to enforce it.
For now, the machinery of war crimes accountability does roughly what its design allows, and no more. It can name perpetrators, build cases that hold up in court and, when a government cooperates, convict and imprison them, as it did with the Darfur commander in October 2025[s]. What it cannot do is reach across a hostile border. The five 2025 withdrawal announcements counted by TRIAL International, one of which Hungary reversed before it took effect[s][s], showed how many governments were willing to wager that the distance between a verdict and a cell is one the Court can never close on its own. The next decade of international justice will be the record of whether they are right.
