KOHEN AVOCATS FIRM · PARIS 16
Nord Stream Croatia Extradition Paused: Why Germany Added a War-Crime Count After Poland Refusal

Hassan KOHEN acts in urgent matters, from the police station to the assize court. Initial strategic analysis: €80 incl. VAT, personal reply within 24 hours.

100% confidential · Legal professional privilege · No obligation

Article generated by artificial intelligence, through a process designed and supervised by the firm

Source: French Court of Cassation – Judilibre and Légifrance open data

Paris Bar Police custody, investigation, assize court CNB listing avocat.fr
Maître Hassan KOHEN, avocat au Barreau de Paris
Hassan KOHEN
Attorney at the Paris Bar

Nord Stream Croatia Extradition Paused: Why Germany Added a War-Crime Count After Poland Refusal

Speak to a lawyer 06 89 11 34 45

On 4 September 2026, a Croatian court was supposed to decide whether Ukrainian national Volodymyr Zhuravlov should be surrendered to Germany over the September 2022 sabotage of the Nord Stream gas pipelines in the Baltic Sea. The hearing never reached the merits. According to his defence team, the German authorities had just updated the European Arrest Warrant with a new legal classification, adding a war-crime charge, and the court postponed the case to give the parties time to address the amended warrant. No new date has been announced. The twist is that Zhuravlov is the same man a Polish court refused to extradite to Germany less than a year earlier: on 17 October 2025, the Regional Court in Warsaw declined to surrender him, lifted his pre-trial detention and, in the words attributed to Polish Prime Minister Donald Tusk, treated the case as closed. He was then rearrested in August 2026 in the Croatian coastal town of Pula on the German warrant, questioned by prosecutors, and placed in detention pending the extradition proceedings, a detention order his lawyer immediately said he would appeal. For an international audience, and especially for anyone who could face a cross-border surrender request while present in France, the Croatian postponement concentrates every hard question of European surrender law into a single file: what happens when the issuing state rewrites its warrant mid-proceedings, whether a second European Arrest Warrant can succeed after a first refusal has become final, how courts test double criminality and the political-offence argument when sabotage is presented as an act of defensive war, and where fundamental-rights defences still bite. This analysis reconstructs the verified sequence from wire reporting and specialised European criminal-law coverage, then explains, through a French lawyer’s lens, exactly what a Paris chambre de l’instruction would test before any surrender from French soil, drawing on the Cour de cassation’s very recent ruling of 29 July 2026 on second warrants after a final refusal. Readers following our extradition coverage may also want our 9 September 2026 briefing on the Tate brothers’ Miami detention ruling and competing extradition requests, which examines detention, flight risk and parallel prosecutions from the same French-procedure perspective.

I. Why the Croatian Court Postponed the Nord Stream Extradition Hearing in September 2026

A. What Germany Changed in the Warrant: the New War-Crime Classification and the Postponed 4 September Hearing

The facts leading to the postponement are recent and largely undisputed. Germany’s Federal Public Prosecutor General, the Generalbundesanwalt, who handles the Nord Stream investigation because the acts allegedly undermined the internal security of the Federal Republic and because of the exceptional importance of the case, has been pursuing seven Ukrainian suspects said to have chartered a yacht and planted explosive devices on the Nord Stream 1 and 2 pipelines in September 2022. The explosions of 26 September 2022 damaged the pipelines so badly that gas could no longer flow from Russia to Europe, and the United States, the United Kingdom and the European Union have all described the blasts as deliberate sabotage. German arrest warrants characterise the conduct as anti-constitutional sabotage committed concurrently with causing an explosion and with destruction of buildings and structures, under Sections 88, 305 and 308 of the German Criminal Code. One suspect track has already run its full course: Ukrainian national Serhij K., treated by the German investigation as a suspected mastermind, was the subject of extended litigation in Italy, where the Court of Cassation first overturned a Bologna surrender authorisation in October 2025 over the executing court’s reclassification of the offence, then confirmed surrender in a second ruling on 19 November 2025, with effective surrender to Germany following on 27 November 2025. On 15 January 2026 the German Federal Court of Justice dismissed his detention complaint, affirming strong suspicion and flight risk, rejecting immunity and combatant-privilege arguments, and holding that Germany has territorial jurisdiction because the result of the offence, the pipelines being rendered inoperable, also occurred on German territory where the pipelines end, as reported in the Court’s press release on case StB 60/25.

The Zhuravlov track is the one now before the Croatian courts, and its most recent turn is what makes this story hot. In August 2026, Croatian police arrested a Ukrainian man in Pula on the German European Arrest Warrant. The Associated Press reported that a Croatian investigative judge then ordered his detention pending extradition proceedings, at the request of prosecutors, and that his lawyer, Matija Milos, would appeal the detention order. German prosecutors described the suspect as a trained scuba diver belonging to the group that planted the explosives near the Danish island of Bornholm, identifying him only as Vladimir Z. A prosecution spokesperson then confirmed to the Associated Press that this was the same man Poland had arrested the previous year and released after the Warsaw court’s refusal to extradite, identified by the Polish authorities as Volodymyr Zhuravlov. The extradition hearing scheduled for 4 September 2026 was therefore set to be the second attempt to surrender the same person on the same German investigation. Then came the amendment: the law firm representing Zhuravlov announced that Germany had updated the warrant by adding a new legal classification of war crime, and the Croatian court postponed the hearing without fixing a new date. The defence framed the postponement as the opening of a new stage in the fight for the rights of a Ukrainian citizen, and specifically questioned whether another European Union country could extradite a Ukrainian national after a Polish court had already refused to hand him over to the same requesting authority.

Three procedural consequences flow from that amendment, and each of them would matter identically before a French court. First, any new classification must be tested against double criminality, meaning the executing judge asks whether the facts described in the warrant would also constitute a criminal offence under the law of the executing state. A war-crime label does not automatically simplify that test: it invites argument about the applicable body of law, the existence of an armed conflict at the relevant time and place, the status of the alleged perpetrators, and whether the domestic criminal code of the executing state captures the same conduct under its own headings. Second, the amendment interacts with the specialty rule, the principle that a surrendered person may only be prosecuted for the offences for which surrender was granted. Defence teams routinely argue that a late-added qualification widens the file after the executing judge’s control, and executing courts answer by examining whether the new description covers the same factual conduct already set out in the warrant or genuinely extends it. Third, the requalification affects the list-based surrender regime: where an offence belongs to one of the serious categories punishable by at least three years in the issuing state, the executing judge surrenders without verifying double criminality and leaves legal characterisation to the issuing authority, so the choice of heading directly determines how deep the executing court’s review goes. The Croatian court’s decision to pause rather than rush past the amended warrant is therefore orthodox rather than surprising: when the issuing state rewrites the legal description, the executing judge must be given the time and the documents to test the new heading against each of these filters.

B. Can Croatia Surrender a Man Poland Already Refused to Hand Over? The Second-Warrant Rule After a Final Refusal

The defence’s deepest argument is not about Croatia at all: it is about Poland. On 17 October 2025, the Regional Court in Warsaw refused to extradite the 46-year-old Ukrainian and lifted his pre-trial detention. As summarised in specialised European criminal-law reporting, the Polish court reasoned first that any action would have fallen within a just defensive war on behalf of Ukraine, so that the suspect had committed no crime under Polish law and double criminality was therefore not established, and second that Germany lacked jurisdiction to prosecute, with jurisdiction instead lying with an international tribunal for armed conflicts at sea or an ad hoc United Nations tribunal. Whether or not one agrees with that reasoning, the refusal became the foundation of the current dispute: the defence now asks how a second executing state can grant what the first executing state definitively refused, and invokes the finality of the Polish decision and the ne bis in idem principle against double jeopardy. This is precisely the question the French Cour de cassation answered only weeks ago, in a decision that any lawyer preparing the Croatian file, or a future French file, must now read closely.

On 29 July 2026, the Criminal Chamber of the Cour de cassation, in decision no. 26-84.072, published in the Bulletin, rejected an appeal against a surrender order that presented a strikingly similar pattern: a Paris chambre de l’instruction had definitively refused surrender to Romania on 29 November 2023, Romania then issued a new European Arrest Warrant for the same sentence on 30 January 2025, and the person was rearrested in France and surrendered under the second warrant. The defence argued that a final refusal vested with res judicata barred re-examination absent a subsequent change of circumstances capable of altering the situation previously recognised by the court, and that ne bis in idem prohibited the second proceedings. The Cour de cassation disagreed on each point, and its reasoning maps directly onto the Zhuravlov configuration. First, endorsing the Court of Justice of the European Union’s ruling of 29 July 2024 in Breian, case C-318/24, the Court held that a final decision refusing surrender rendered by the executing judicial authority of a Member State does not deprive the issuing judicial authority of the possibility of maintaining its surrender request, provided the conditions required by European Union law are met. Second, and most quotably, the Court held that the November 2023 refusal concerned a distinct European Arrest Warrant, so that, for lack of an identical object, its authority cannot be opposed to the examination of the warrant at issue in the present proceedings. Third, the Court stated that proceedings for the execution of a European Arrest Warrant issued for the enforcement of a conviction do not constitute a new prosecution for the facts underlying that conviction, so that the ne bis in idem principle does not apply to them. The appeal was therefore rejected and surrender authorised despite the earlier final refusal.

Applied to Croatia, this framework suggests the Polish refusal is a serious argument but not an automatic shield. The Croatian court will have to determine whether the German warrant now before it is, legally speaking, the same object as the one Warsaw rejected: the addition of the war-crime classification cuts both ways here, since the defence can present it as an alteration confirming that the file keeps shifting, while the requesting authority can present it as the very new element that justifies re-examination. The Court of Justice’s Breian logic, as received in French law, means the issuing state is entitled to maintain and even renew its request, and each executing state examines the warrant before it on its own grounds, under its own double-criminality and fundamental-rights review. A refusal in Warsaw does not travel as res judicata to Pula or Zagreb, and it would not travel to Paris either: had Zhuravlov been arrested in France instead of Croatia, the French chambre de l’instruction would examine the German warrant afresh, testing the new war-crime heading against French double criminality, the mandatory and optional refusal grounds of the Code of Criminal Procedure, and fundamental rights, without treating the Polish decision as binding. That is the first lesson of the French lens: in European surrender law, finality protects the person against repeated prosecution for the same facts, but it does not freeze the requesting state’s file forever, and each new warrant before each new executing judge restarts the control from zero.

II. What France Would Test Before Surrendering a Nord Stream Suspect

A. Double Criminality, Political Offence and the Terrorism Exception Under French Law

If the German Nord Stream warrant ever reached a French court, the examination would begin with the architecture of the French surrender system. The European Arrest Warrant, defined by Article 695-11 of the Code of Criminal Procedure as a judicial decision issued by one Member State for arrest and surrender by another, for prosecution or for enforcement of a custodial sentence, is handled entirely by the judiciary, unlike classic extradition to non-European Union states, where the chambre de l’instruction gives only a reasoned opinion and the government decides by decree. The general extradition chapter still frames the philosophy: Article 696 provides that, absent an international convention providing otherwise, the conditions, procedure and effects of extradition are determined by the provisions of that chapter. Within classic extradition, Article 696-4 lists the cases where extradition is not granted, and its second ground is worded in terms any Nord Stream defence would immediately seize: extradition is refused where the crime or offence is of a political character, or where the circumstances show that extradition is requested for a political purpose. The provision continues with the other classic bars, including offences committed on French territory and cases already finally tried in France. A sabotage presented by the defence as an act of defensive war in an international armed conflict would therefore be argued, in French classic-extradition terms, as falling squarely within the political-offence exception, exactly as the Warsaw court reasoned under Polish law.

Inside the European Arrest Warrant system, however, the political-offence exception as such does not exist: the Framework Decision 2002/584/JHA of 13 June 2002 deliberately removed it to build mutual recognition, and French law transposes that choice faithfully. The French judge therefore converts the political-defence argument into neighbouring controls. The first is double criminality under Article 695-23, which allows refusal where the facts in the warrant do not constitute an offence under French law, while providing that the warrant is executed without any double-criminality check where the conduct is punishable in the issuing state by at least three years of imprisonment and falls within one of the serious categories listed in Article 694-32, a list that includes terrorism, trafficking in arms, munitions and explosives, sabotage-adjacent headings and participation in a criminal organisation. Legal characterisation and the applicable penalty are then for the issuing authority alone to assess. This is where Germany’s choice of headings becomes decisive for any French control: sabotage charged as terrorism or as an explosives-trafficking category offence punishable by three years or more would travel to Paris without double-criminality review, whereas a novel or contested war-crime characterisation would invite the French court to verify precisely how French law captures underwater pipeline sabotage committed outside France, including French provisions on destruction, explosives and, where applicable, terrorism. The requesting state’s amendment of the warrant is thus never neutral: each new heading redraws the boundary between what the executing judge must verify and what it must take on trust.

The second conversion runs through the refusal grounds of Article 695-22 and Article 695-24. Article 695-22 imposes mandatory refusal in defined cases, including where the person has already been finally judged for the same facts by the French authorities or by another Member State than the issuing state, provided that in case of conviction the sentence has been served, is being served, or can no longer be executed. Article 695-24 adds optional grounds, including where French courts are already prosecuting the same facts or have decided not to prosecute or to discontinue, and where the offence was committed in whole or in part on French territory. None of these grounds reproduces the Warsaw reasoning directly, which is the point: a French court could not refuse surrender on the ground that the sabotage was a legitimate act of war, because that justification does not appear among the statutory refusal grounds, and the German Federal Court of Justice has already rejected the cognate immunity and combatant-privilege arguments on the merits in the Serhij K. proceedings. What the French court can do, and would do, is test whether the same facts have already been finally adjudicated elsewhere, whether France itself claims jurisdiction over part of the conduct, and whether the file demonstrates the precise correspondence between the facts described and the offence headings invoked. The Polish court’s jurisdictional objection, that only an international or ad hoc tribunal could judge the Bornholm explosions, would likewise find no direct equivalent in the French refusal catalogue: the French judge examines its own statutory grounds, not the issuing state’s jurisdiction in the abstract, and Germany’s ubiquity-style reasoning, locating the result of the offence on German territory where the pipelines terminate, would be taken as the issuing state’s characterisation rather than relitigated as a preliminary plea.

B. Detention, Fundamental Rights and the Final French Filter Before Any Surrender

Procedure matters as much as substance, and here the Croatian sequence already mirrors what French practice would produce. After notification of a European Arrest Warrant, where the prosecutor does not leave the person at liberty, Article 695-28 requires presentation to the first president of the cour d’appel, who orders committal to the remand prison of the court’s seat unless satisfied that the person’s appearance at every stage is sufficiently guaranteed, with supervision measures available as the alternative. The Croatian detention pending extradition proceedings, and the immediate appeal against it, therefore reflect the standard European pattern: in surrender cases involving foreign nationals, prior cross-border flight and grave allegations, detention is the default and release the exception requiring concrete guarantees. The same logic governed the German detention ruling of January 2026, where flight risk anchored continued custody, and it governs French chambers daily. Defence teams should therefore prepare the detention file first, before the merits: stable address, employment, family ties, passport surrender, electronic monitoring proposal and a concrete plan for appearing at each hearing. Waiting for the surrender hearing to address detention is a strategic error, because the custody decision shapes everything that follows, from access to the file to the practical ability to instruct experts on foreign law.

The hearing itself would then run through two further French filters. The first is formal regularity. The 29 July 2026 Cour de cassation decision is instructive well beyond its ne bis in idem holding: the Court recalled that questioning of the requested person by the chambre de l’instruction under Article 695-30 aims only at establishing identity, receiving observations on the proceedings and allowing consent or refusal, so that this formality is preliminary to any debate, while the reporting judge’s report must precede the debate itself, with both formalities validly accomplished before argument begins. Defence challenges based on the order of these formalities will therefore fail where both were completed before the debate, and the lesson for practitioners is to aim procedural challenges at genuine irregularities, missing documents, imprecise facts or dates in the warrant, absence of translation, rather than at the choreography of the hearing. The second filter is fundamental rights under Article 1(3) of the Framework Decision, which the Cour de cassation in the same July 2026 ruling confirmed the lower court had examined and set aside through reasoning distinct from the grounds criticised on appeal. French courts take these pleas seriously but demand proof of a real, individualised risk, not general denunciations of the requesting state’s justice system. The standard was restated the previous year in decision no. 24-85.773 of 6 May 2025, concerning extradition requested by South Korea, where the Criminal Chamber recalled that extradition must be refused where the person shows exposure to a real risk of a manifestly disproportionate sentence in the requesting state, a notion subject to strict interpretation and sustainable only in very exceptional cases, citing the European Court of Human Rights ruling of 17 January 2012 in Harkins and Edwards v. the United Kingdom. Transposed to a Nord Stream file, this means arguments about sentence severity, detention conditions or the political context of the German prosecution would need to be documented person by person and count by count, with concrete evidence rather than press narratives, and measured against the reality that Germany is a European Union Member State with enforceable judicial remedies, which raises the evidentiary bar considerably.

Conclusion

The postponed Croatian hearing is a snapshot of European surrender law under stress, and its outcome will be read far beyond Zagreb. Germany, having secured one surrender through Italy and a detention ruling at home, is now pressing its second warrant against a man Poland refused to hand over, and has just raised the stakes by adding a war-crime classification that forced the Croatian court to pause. The defence, having won in Warsaw on defensive-war and jurisdictional grounds, is now testing whether finality, double criminality and the specialty rule can stop the second attempt in Croatia. The French reference points assembled here suggest how that test should be structured wherever it is litigated: a final refusal in one executing state does not bind the next executing judge, because each warrant is examined on its own object and surrender proceedings are not a new prosecution for ne bis in idem purposes, as the Cour de cassation confirmed on 29 July 2026; each new legal heading redraws the double-criminality boundary and must be verified against the executing state’s own law unless it falls within the serious list-based categories; the political or defensive-war justification, decisive in Warsaw, has no direct equivalent among French European Arrest Warrant refusal grounds and would have to be converted into double criminality, prior-final-decision or fundamental-rights arguments; and detention, formal regularity and individualised rights evidence will decide the case as much as grand theory. For companies, insurers, journalists and families following the Nord Stream saga, the practical message is that surrender litigation is won file by file, heading by heading and hearing by hearing. For any person exposed to a cross-border warrant while in France, the message is more urgent: the French chambre de l’instruction will examine the warrant afresh, without deference to Warsaw or Zagreb, and the weeks between arrest and hearing are the only window in which double criminality, specialty, limitation, translation and guarantees can be litigated effectively. Use them.

📄 Official circular