On 24 September 2026, a county court in Pula, Croatia, approved the extradition to Germany of Volodymyr Zhuravlov, the Ukrainian scuba diver whom Berlin suspects of helping blow up the Nord Stream gas pipelines in September 2022. His lawyer denounced a biased and political ruling and announced an appeal, which suspends any transfer until the appellate court decides. Four years after the explosions off the Danish island of Bornholm, the investigation has therefore entered its most delicate phase: not the trial itself, but the cross-border machinery that decides who stands trial where. That machinery is the European Arrest Warrant, the fast-track surrender system that replaced classic extradition between European Union member states and now moves thousands of suspects across borders every year. Croatia is in the Union, Germany is in the Union, and the warrant that put Zhuravlov in a Pula courtroom in August 2026 travels under rules written in Brussels and interpreted in Luxembourg, not under bilateral diplomacy. For an international reader, the case raises three blunt questions. First, on what basis does Germany claim the right to prosecute a sabotage committed in international waters near Denmark? Second, what defences can a requested person realistically raise once a European Arrest Warrant lands, given that the system was designed to make surrender almost automatic? Third, how would the same file be handled if the arrest had happened in Paris rather than Pula? France applies the same European rules, but its Cour de cassation has spent the summer of 2026 drawing sharp lines around refusal grounds, family of case law this article examines through three rulings handed down in February and July 2026. The French approach matters beyond France: defence lawyers across Europe cite it, and anyone facing surrender proceedings in any member state will meet the same tests of mutual trust, fundamental rights, double jeopardy and specialty. This analysis reconstructs the Pula decision from the wire reports, explains Germany’s jurisdictional claim, and then walks through the French checklist that determines whether a European Arrest Warrant ends in a boarding gate or in a refusal.
I. A Croatian court says yes to Germany: the Nord Stream file and the jurisdictional bet behind it
A. What the Pula court decided about Volodymyr Zhuravlov and what happens next
Volodymyr Zhuravlov, rendered in the German press as Vladimir Z., is a Ukrainian national described by German prosecutors as a recreational-turned-operational scuba diver who formed part of the group that planted explosives on the Nord Stream 1 and Nord Stream 2 pipelines in September 2022. Croatian police picked him up in August 2026 in a hotel in the coastal city of Pula, where Ukrainian media report he was working on a film shoot, on the strength of a European Arrest Warrant issued by the German federal prosecution. He denies any involvement in the sabotage, a position his lawyers have maintained since the first hearing. On 20 August 2026 he was escorted by police to a hearing before the County Court of Pula, and the court then took the surrender question under advisement. The ruling itself was made on Friday 18 September 2026 and publicly confirmed on Thursday 24 September: the court authorised his extradition, in European Union vocabulary his surrender, to Germany. The court declined to give reasons in public until the parties decide on an appeal, which is standard practice in surrender litigation where detailed reasoning could prejudice parallel investigations. Zhuravlov’s lawyer, Mykola Katerynchuk, immediately called the ruling biased and political and announced an appeal, which under Croatian surrender procedure keeps his client in Croatia pending the higher court’s review. That appeal is now the whole case in miniature, because the appellate judges will have to answer the same questions every executing court in Europe must answer when a European Arrest Warrant arrives: is the warrant formally valid, does any mandatory ground for refusal apply, and has the defence raised a credible, evidence-backed risk to fundamental rights that the issuing state has failed to dispel. Nothing in the wire reports suggests the defence disputes the identity of the person or the formal regularity of the German warrant, so the fight will concentrate on the substance, on Germany’s underlying accusation and on whether surrender exposes a Ukrainian national, arrested in a Union member state for acts allegedly committed in the service of Ukraine’s war effort, to a prosecution his lawyers already frame as political. Readers should keep one procedural fact in mind: within the European Union, surrender is not extradition. There is no ministerial decree, no diplomatic negotiation, and no executive discretion of the kind that still governs extradition to or from third states. A judge in Pula applies European rules, reviews the file, hears the person, and authorises or refuses the transfer, subject only to appeal within the judiciary. That judicialisation is precisely what makes the coming Croatian appellate ruling worth watching for lawyers far beyond Zagreb, because the arguments raised there, political motivation, evidentiary thinness, proportionality, are the same ones raised every week in Paris, Madrid and Warsaw.
B. Why Germany claims jurisdiction over explosions off Bornholm and who else is in the frame
Germany’s claim to prosecute rests on two pillars that the wire reports lay out plainly. First, geography and infrastructure: the damaged pipelines come ashore in the German state of Mecklenburg-Vorpommern, and German courts treat the case as falling within their jurisdiction because the destruction of infrastructure terminating on German soil struck directly at Germany’s energy security and internal safety. Under the law of the sea and under German criminal law, a state may assert jurisdiction where the effects of an attack materialise on its territory, and the sudden loss of the Baltic gas route in the autumn of 2022 was, for Germany, an attack on a vital supply artery only months after Russia’s full-scale invasion of Ukraine had turned energy into a weapon. Second, investigative attribution: after years in which Swedish, Danish and German inquiries ran in parallel and Stockholm and Copenhagen eventually closed their files, the German federal prosecution kept going, identified a sailing boat allegedly used by the sabotage team, and traced individual suspects through diving equipment, travel records and witness accounts. Zhuravlov is not the only name on the German list. In July 2026, federal prosecutors charged a former Ukrainian army officer identified as Serhii K. with orchestrating the blasts while acting on behalf of Ukrainian state entities, a formulation that, if proven, would transform the case from freelance sabotage into a state-sponsored covert operation against allied infrastructure. Serhii K. was arrested in Italy in August 2025 and later transferred to Germany, where his case is moving toward what could become the first full trial of the Nord Stream affair, reportedly in Hamburg. He too denies any involvement. The coexistence of the two files explains why the Pula ruling matters beyond one diver’s fate: if Zhuravlov is surrendered, German prosecutors would hold two alleged team members on German soil and could join or sequence the proceedings, with each defendant’s statements potentially affecting the other’s trial. For the defence, the state-entity allegation against Serhii K. is double-edged. It feeds the narrative that the prosecution is political, a Ukrainian acting against Russian infrastructure at a time when Russian gas revenues funded the invasion of his country, yet within the European Arrest Warrant system the political character of an offence is largely irrelevant between member states, a point this article develops below through French law. The Ukrainian government’s position adds another layer: Kyiv has consistently denied official involvement in the sabotage, and any trial in Germany that tests the claim of state sponsorship would carry diplomatic consequences well beyond the courtroom. That is the backdrop against which the Croatian appellate court will rule in the coming weeks, and against which French judges would rule if the same warrant ever surfaced at Roissy or in a Paris police station.
II. Surrender is the rule, refusal the exception: the French checklist every European Arrest Warrant must survive
A. Mutual trust first, then the closed list of refusal grounds: double jeopardy and double criminality under the Cour de cassation’s summer case law
The starting point of French law is brutally simple and it is the same starting point a Croatian judge applies: within the European Union, a judicial decision from one member state is recognised and executed in the others almost automatically. The Cour de cassation restated the principle on 29 July 2026 in a ruling quashing a Paris decision that had refused to surrender a woman to Germany, holding that “l’exécution d’un mandat d’arrêt européen ne saurait être refusée pour des motifs autres que ceux que prévoient la décision-cadre et les textes pris pour son application” (Cass. crim., 29 July 2026, no. 26-84.256). In plain English: a European Arrest Warrant cannot be refused on any ground other than those provided by the Framework Decision and the statutes that transpose it. The Framework Decision in question is Council Framework Decision 2002/584/JHA of 13 June 2002, available on the EUR-Lex public portal, and its French transposition lives in the Code of Criminal Procedure. That closed-list logic is the first thing an international reader must grasp, because it reverses the reflexes built by classic extradition: the executing judge does not ask whether prosecution in the issuing state seems wise, fair or well-evidenced, but only whether one of the exhaustively listed refusal grounds is made out. French statute organises those grounds in two tiers. Article 695-22 of the Code of Criminal Procedure lists the mandatory refusals: amnesty in France covering the acts, a final judgment for the same acts already handed down in France or in another member state, criminal irresponsibility by reason of age, and, crucially, proof that the warrant was issued to prosecute or punish a person on account of sex, race, religion, ethnic origin, nationality, language, political opinions, sexual orientation or gender identity, or that surrender would harm the person for one of those reasons. That fifth paragraph is the only door through which an argument of political persecution can enter European Arrest Warrant litigation, and the Cour de cassation treats it as a narrow, evidence-driven exception rather than a general political-offence clause. The narrower scope is deliberate: unlike classic extradition, where article 696-4 of the Code of Criminal Procedure provides that “L’extradition n’est pas accordée” where the offence has a political character or the request is made for a political purpose, the European system abolished the political-offence exception between member states and replaced it with the discrimination test just described. A lawyer who walks into a Paris surrender hearing arguing, as Zhuravlov’s counsel did in Pula, that the prosecution is political must therefore translate that claim into the discrimination vocabulary of article 695-22 or lose it. The second tier is optional refusal, and its flagship is double criminality: article 695-23 of the Code of Criminal Procedure states that “L’exécution d’un mandat d’arrêt européen peut également être refusée si le fait faisant l’objet dudit mandat d’arrêt ne constitue pas une infraction au regard de la loi française”, while waiving any double-criminality check for a list of serious offences punishable by at least three years where they fall into the categories of article 694-32. Sabotage of energy infrastructure, terrorism-related charges and organised crime all sit comfortably inside that waiver, which means a French court would almost certainly decline to second-guess Germany’s characterisation of pipeline sabotage as a crime. The heaviest litigation in Paris this year has instead concerned the ne bis in idem ground, the rule that no one is tried twice for the same facts. On 16 July 2026 the Cour de cassation quashed a Paris ruling that had authorised the deferred surrender of an alleged former ETA leader to Spain, because the lower court had failed to verify precisely whether the Spanish warrant covered the same material facts as a French acquittal and ongoing French proceedings (Cass. crim., 16 July 2026, no. 26-83.852). The ruling recalls that the concept of the same facts is autonomous in Union law and must “être interprétée comme visant la seule matérialité des faits et englobant un ensemble de circonstances concrètes indissociablement liées entre elles, indépendamment de la qualification juridique de ces faits ou de l’intérêt juridique protégé”, following the Court of Justice’s Mantello judgment (CJEU, Grand Chamber, 16 Nov. 2010, Mantello, C-261/09). It then faults the Paris court for ruling on a warrant that merely stated the person had acted as a leading ETA figure between 2002 and 2019, “sans autre indication concernant la matérialité des faits, leur date, leur lieu et les circonstances concrètes de leur commission de nature à lui permettre de statuer sur les motifs de refus pertinents”. The lesson for surrender litigation everywhere in Europe is procedural and goes straight to the Zhuravlov file: where the issuing state’s description of the facts is vague about dates, places and concrete circumstances, the executing court must demand supplementary information instead of guessing, and a refusal to ask can itself justify quashing the surrender. Applied to Pula, the question is whether Germany’s warrant describes Zhuravlov’s alleged dives with enough material precision, dates, boat, role in the team, for the Croatian court to exclude double jeopardy and verify double criminality, and whether the appellate court will press Berlin for more detail before letting the transfer proceed.
B. The three defences that actually work in Paris: health and fundamental rights, specialty, and strict procedure before the chambre de l’instruction
Once the closed list is understood, experience before the Paris chambre de l’instruction shows that only three lines of defence regularly succeed, and each is tightly framed by recent case law. The first is the fundamental-rights objection based on health or detention conditions. The 29 July 2026 ruling adds a demanding two-stage test borrowed from the Court of Justice’s E.D.L. judgment (CJEU, 18 Apr. 2023, E.D.L., C-699/21): where the executing court finds serious and proven grounds to believe surrender would expose a gravely ill person to a real risk of a significant reduction in life expectancy or a rapid, significant and irreversible deterioration of health, it must suspend surrender and ask the issuing state how detention and care would be adapted, and it may refuse surrender only if the information gathered shows the risk cannot be ruled out within a reasonable time. The Cour de cassation quashed the Paris refusal precisely because the lower court had rejected German medical assurances as too general without establishing “que la remise de la personne recherchée exposerait cette dernière à un risque réel de réduction significative de son espérance de vie ou de détérioration rapide, significative et irrémédiable de son état de santé ne pouvant pas être écarté dans un délai raisonnable” (Cass. crim., 29 July 2026, no. 26-84.256, para. 17). For Zhuravlov, whose file as reported contains no medical dimension, this door appears closed, but the ruling illustrates the broader method: bare assertions never suffice, the defence must document a real, individualised risk with expert evidence, and the issuing state’s guarantees are then tested rather than dismissed. The second working defence is the specialty rule, the guarantee that a surrendered person will only face the charges named in the warrant. On 10 February 2026 the Cour de cassation recalled that “en présence d’une infraction autre que celle qui a motivé la remise, la personne remise ne saurait être poursuivie, condamnée ou privée de liberté sans son consentement, recueilli dans les formes prescrites” (Cass. crim., 10 Feb. 2026, no. 25-87.769), drawing on the Court of Justice’s Leymann ruling, and restated the statutory core that a person surrendered to France “ne peut être poursuivie, condamnée ou détenue en vue de l’exécution d’une peine privative de liberté pour un fait quelconque antérieur à la remise et autre que celui qui a motivé cette mesure, sauf dans l’un des cas énumérés audit article”, namely article 695-18 of the Code of Criminal Procedure. In the Nord Stream context, specialty is far from academic: if Zhuravlov were surrendered for sabotage and German investigators then wished to prosecute him for espionage, membership of a foreign intelligence operation, or any earlier conduct, they would need his consent or a fresh authorisation from the executing state. Defence teams in multi-suspect files monitor this closely, because statements made by one surrendered suspect can generate new charges against another, and each extension of the prosecution beyond the warrant reopens judicial control. The French hearing itself provides the third line of defence, pure procedure. Article 695-29 of the Code of Criminal Procedure requires that “La chambre de l’instruction est immédiatement saisie de la procédure” and that the wanted person appear within five working days of presentation to the prosecutor general, while article 695-30 organises the adversarial hearing with counsel and, where needed, an interpreter, and article 695-31 frames consent to surrender and the separate, irrevocable waiver of specialty. The Cour de cassation polices these requirements strictly: insufficient or contradictory reasoning equals no reasoning, and any failure to answer the essential arguments in the parties’ briefs sends the case back. The 16 July 2026 ETA ruling demonstrates the sanction in action, with a partial quashing limited to the surrender order and a remand before differently composed judges. Transposed to the Adriatic, the message to the Croatian appellate court is identical in spirit: answer every essential defence argument, on the facts as precisely described, or face reversal. None of this gives Zhuravlov an easy victory. The European Arrest Warrant was built to deliver people like him, a suspect named in a formally valid warrant for grave offences carrying heavy penalties, with no prior final judgment on the same facts and no documented medical risk. His realistic hopes lie in the margins the French case law maps so carefully: forcing Berlin to particularise the alleged dives, testing whether any parallel Ukrainian or third-state proceedings trigger ne bis in idem, and preserving specialty so that a sabotage surrender cannot quietly become an espionage trial. Those margins are narrow, but in surrender litigation narrow margins are the entire profession.
Conclusion
The Pula ruling of 24 September 2026 does not decide whether Volodymyr Zhuravlov blew up anything; it decides that a Croatian judge finds Germany’s paperwork sufficient to let German judges ask that question at trial. That distinction defines the whole European Arrest Warrant era: executing courts police the transfer, not the merits, and they do so through a closed list of refusal grounds interpreted identically from Zagreb to Paris. Germany’s jurisdictional bet, that sabotage in the Baltic is prosecutable where its effects crippled German energy security, will now be tested twice, first on appeal in Croatia on surrender law, then, if the transfer goes through, at trial on criminal evidence. The French rulings of February and July 2026 give the clearest available map of the first test: surrender cannot be refused for reasons outside the Framework Decision, the same facts must be compared by their materiality rather than their legal labels, vague warrants must trigger requests for detail rather than benefit-of-the-doubt surrenders, health objections must prove a real and irremediable risk, and specialty fences the prosecution to the charges named in the warrant. For states, the Nord Stream files signal that covert attacks on allied infrastructure will be pursued across borders for years, through successive warrants, until every alleged team member faces a courtroom. For individuals, the signal is symmetrical: once a European Arrest Warrant exists, only rigorous, documented, procedurally exact defence work can stop it, and the earlier that work starts, ideally from the first appearance before the executing judge, the wider the remaining margins. Anyone arrested on a European warrant, in France or elsewhere, should therefore do three things at once: demand the precise material facts behind each charge, assemble evidence for any fundamental-rights risk without delay, and lock down specialty before consenting to anything. The Croatian appellate court’s coming decision will show whether Zhuravlov’s team has done exactly that.