On 24 September 2026, at 9:08 AM Eastern time, the Reuters wire moved a short dispatch that every extradition lawyer in Europe read twice. A Croatian court, the County Court in Pula, had ruled that a Ukrainian man suspected of helping to blow up the Nord Stream gas pipelines in September 2022 can be extradited to Germany. The man, identified by German prosecutors only as Vladimir Z., a trained scuba diver, was arrested on 20 August in a hotel in the coastal city of Pula on a German European arrest warrant. He denies any involvement. His lawyer, Mykola Katerynchuk, dismissed the court ruling, made on Friday last week, as biased and political and said there would be an appeal, according to the Suspilne report relayed by Reuters. The County Court in Pula confirmed on Thursday it had made a ruling, but said it could not go into details until the parties had decided whether to appeal. Why does this ruling matter far beyond Croatia? Because barely a year earlier, another European court, in Poland, looked at the same German case against the same network of suspects and answered the exact opposite: no surrender. And because the legal machinery that produced both answers, the European arrest warrant, is the same instrument a Paris chambre de l’instruction applies every week. This article reconstructs what Berlin charges, explains why Warsaw and Pula diverged, and sets out precisely what would happen if the diver had been arrested in France instead of Croatia.
I. Can a Ukrainian Nord Stream sabotage suspect be extradited to Germany?
A. What charges does the German European arrest warrant contain?
The German prosecution case, as reported by the BBC on 19 August 2026 and confirmed by the federal prosecutor’s office, is strikingly concrete. The suspect is alleged by German authorities to have been one of the scuba divers who planted explosives on the pipelines in September 2022. In the words of the German prosecutors relayed that day, the suspect participated in the necessary dives for the sabotage, and once surrendered he will be brought before an investigating judge to face charges of causing explosions and sabotage. Three of the four Nord Stream pipelines were ruptured by the underwater explosions of September 2022, an attack that crippled a long-controversial energy supply line from Russia to Germany and that both Russia and Western countries described as sabotage. German courts have treated the case as falling within their jurisdiction because the damaged pipelines terminate in the German state of Mecklenburg-Vorpommern, and their destruction affected Germany’s energy security and internal safety. That jurisdictional reasoning matters for everything that follows: Germany does not claim universal jurisdiction over the Baltic seabed. It claims that the effects of the explosions struck German territory, German energy supply and German internal safety, which under German law founds a genuine territorial link for prosecution. For an extradition lawyer, this is the first checkpoint of any surrender file: the issuing state must show that it is prosecuting an offence over which it has jurisdiction, and that the factual description in the warrant is precise enough for the executing court to test the refusal grounds. Precision is not a formality. As the French Cour de cassation recalled only weeks ago, in a ruling of 16 July 2026 discussed below, a European arrest warrant that describes years of alleged conduct without materiality, dates, places or concrete circumstances does not allow the executing court to rule on the refusal grounds, and the surrender must be annulled. The German warrant against the diver, by contrast, names dives, explosives, pipelines and a precise month, which is exactly the kind of factual density that survives judicial review.
The diver is not the first suspect in this investigation, and the comparison with the parallel file is instructive. Prosecutors allege that a former Ukrainian military officer, Serhii K., led a team that used a yacht rented at Rostock in Germany to plant explosives on the pipelines close to the Danish Baltic Sea island of Bornholm. That man was arrested in Italy in August 2025 and transferred to Germany in November, appeared before a court in Hamburg in June 2026, and denies responsibility. Then, according to Reuters, that same suspect was charged by prosecutors in July with orchestrating the blasts while acting on behalf of Ukrainian state entities. The German investigators had already told the United Nations Security Council, in a letter reported by Reuters at the time of the explosions, that traces of subsea explosives were found in the samples taken from the boat. Two consequences flow from this for the surrender analysis. First, Italy surrendered its suspect to Germany without the kind of judicial resistance seen in Poland, which shows that within the European arrest warrant system the default is surrender and refusal is the exception. Second, the allegation that the operation was conducted on behalf of Ukrainian state entities injects a state-action dimension into the file: defence teams will argue that a sabotage operation attributed to a state at war belongs to a different legal universe than ordinary organised crime, while issuing authorities will answer that motive does not erase the criminal nature of planting military-grade explosives under civilian energy infrastructure. That clash already decided one surrender case, in Warsaw, and it is now heading for an appeal in Croatia.
B. Why did Poland refuse surrender while Croatia authorised it?
The Polish episode remains the most commented surrender refusal in recent European practice. The same network’s alleged diver, understood to be Volodymyr Zhuravlyov, had been arrested in Poland in 2025 on the German warrant. As the BBC reported, a Polish judge refused a German extradition request, arguing that if Ukraine was responsible for the attack, then it was a just act. He was then released, prompting sharp criticism from Berlin. Put in doctrinal terms, the Polish court introduced a substantive political-legitimacy test into a procedure that, under European Union law, does not contain one. The Council Framework Decision of 13 June 2002 on the European arrest warrant and the surrender procedures between Member States, available on the official EUR-Lex database at Framework Decision 2002/584/JHA, organises surrender on the principle of mutual recognition, and its articles 3, 4 and 4a list exhaustively the grounds on which an executing court may or must refuse. A finding that the alleged sabotage was justified by the context of the war in Ukraine appears nowhere in that list. The Croatian court in Pula has now taken the opposite route: it verified the warrant, found no applicable refusal ground, and authorised surrender, leaving the political argument to the appellate court that will hear the announced appeal. Two executing courts, one warrant system, two answers. For practitioners, the lesson is that the European arrest warrant leaves the assessment of opportunity and proportionality to the issuing authority alone, and an executing judge who substitutes a personal view of the justice of the cause steps outside the instrument.
French case law states that boundary with unusual clarity, and it does so in two rulings handed down within weeks of the Nord Stream decisions. On 6 August 2025, the criminal chamber of the Cour de cassation, in decision number 25-84.847 published at Cass. crim., 6 August 2025, no. 25-84.847, quashed a ruling by the Paris chambre de l’instruction that had refused to surrender a convicted man to Romania. The Paris judges had relied on a bundle of considerations: the age of the facts, the state of health of the person concerned, a parallel civil dispute with the Romanian state, the withdrawal of an Interpol notice, and prior refusals in other countries after findings of unfair trial. The Cour de cassation swept that reasoning aside in terms that deserve full quotation: “Sous réserve du respect des droits fondamentaux de la personne recherchée et des principes juridiques fondamentaux consacrés par l’article 6 du Traité sur l’Union européenne, garanti par l’article 1, § 3, de la décision-cadre 2002/584/JAI du Conseil, du 13 juin 2002, relative au mandat d’arrêt européen et aux procédures de remise entre Etats membres, telle que modifiée par la décision-cadre 2009/299/JAI du Conseil du 26 février 2009, 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.” The holding continues, even more directly on the point that sank the Polish-style reasoning: “elle ne pouvait refuser la remise de M. [K] au regard du caractère disproportionné de la délivrance du mandat d’arrêt européen, qui ne figure pas parmi les motifs de non-exécution obligatoire ou facultative, prévus aux articles 3, 4 et 4 bis de la décision-cadre 2002/584/JAI, cette appréciation relevant de la seule autorité de l’Etat d’émission.” In other words, whether issuing the warrant was proportionate or politically opportune is the issuing state’s business, not the executing court’s. A Pula judge who authorises surrender despite the war context, and a Warsaw judge who refuses it because of that same context, are not two equally valid readings of the instrument: under the mutual-recognition logic recalled by Paris, only the first stays inside the Framework Decision. The announced Croatian appeal will therefore have to be fought on the closed list of refusal grounds, fundamental rights evidence included, and not on the justice of the Ukrainian cause, however sympathetic a court may find it.
II. What would happen if the suspect had been arrested in France?
A. How does the Paris chambre de l’instruction decide on surrender?
Imagine the diver arrested not in a Pula hotel but in a Paris one, on the same German European arrest warrant. The procedure that would follow is entirely judicial, entirely codified, and far faster than ordinary extradition. After arrest, the person is brought before the prosecutor general, who notifies the warrant, and the file goes to the chambre de l’instruction of the Paris court of appeal, the specialised fifth section that handled both of the 2025 and 2026 surrender cases discussed in this article. The first question the chamber asks is whether the person consents to surrender, after being informed of the legal consequences of consent and of its irrevocable character. The Code of criminal procedure, in article 695-31 of the Code de procédure pénale, in force as verified on 24 September 2026, provides: “Si la chambre de l’instruction constate que les conditions légales d’exécution du mandat d’arrêt européen sont remplies, elle rend un arrêt par lequel elle donne acte à la personne recherchée de son consentement à être remise ainsi que, le cas échéant, de sa renonciation à la règle de la spécialité et accorde la remise.” The same article sets the tempo: “La chambre de l’instruction statue, sauf si un complément d’information a été ordonné dans les conditions énoncées à l’article 695-33, dans les sept jours de la comparution devant elle de la personne recherchée. Cette décision n’est pas susceptible de recours.” Seven days, no appeal: consenting to surrender ends the case in a week, and the consent cannot be taken back. That is why the first advice any Paris extradition lawyer gives a client arrested on a European warrant is to say nothing about consent before the file has been read in full, because consent also opens the question of waiving the specialty rule, the protection that prevents the issuing state from prosecuting the surrendered person for offences other than those listed in the warrant. The chamber must ask the person whether they intend to waive specialty, after informing them of the legal consequences of such a waiver and of its irrevocable character. In the Nord Stream file, that waiver would matter enormously: a surrender for causing explosions and sabotage, without a specialty waiver, would bar Germany from adding unrelated charges later without French authorisation.
If the person does not consent, and the diver in Pula has denied everything so he would certainly not consent in Paris either, the same article 695-31 of the Code de procédure pénale continues: “Si la personne recherchée déclare ne pas consentir à sa remise, la chambre de l’instruction statue par une décision dans le délai de vingt jours à compter de la date de sa comparution” (article 695-31 of the Code de procédure pénale). That decision, unlike the consent order, can be challenged by a pourvoi en cassation, by the prosecutor general or by the person concerned. In practice the person is often placed under judicial supervision while the chamber examines the file, exactly as happened in the July 2026 case where the Spanish warrant of 3 December 2025 was notified on 11 December 2025 and the person was placed under contrôle judiciaire. The chamber then tests each refusal ground in adversarial debate, with defence memoranda and the prosecution’s submissions, and it must answer the essential arguments of the parties: insufficient or contradictory reasoning is treated as no reasoning at all, and the Cour de cassation quashes on that basis alone. Where the warrant is too vague to allow that review, the chamber does not guess and does not refuse outright: it asks the issuing authority for more. The Cour de cassation recalled the mechanism in its ruling of 16 July 2026, number 26-83.852, published at Cass. crim., 16 July 2026, no. 26-83.852, in these terms: “si la chambre de l’instruction estime que les informations communiquées par l’État membre d’émission dans le mandat d’arrêt européen sont insuffisantes pour lui permettre de statuer sur la remise, elle demande à l’autorité judiciaire dudit État la fourniture, dans le délai maximum de dix jours pour leur réception, des informations complémentaires nécessaires.” Ten days maximum for the reply. The Paris procedure is therefore a high-speed adversarial audit of the warrant, not a retrial of the facts, and the German file on the Nord Stream dives, with its named dives, named boat, named island and named month, is precisely the kind of detailed warrant that passes such an audit. The vapour-thin warrant that the Cour de cassation struck down in July 2026, which accused a man of having acted as one of the principal leaders of ETA over seventeen years without a single material fact, date, place or concrete circumstance, is the counter-model, and it shows what a defence team in Paris would attack first: not the war, not the politics, but the paper.
B. On which legal grounds could a French court refuse surrender?
French law closes the list of refusal grounds exactly as the Framework Decision does, and it divides them into mandatory and optional bars. The mandatory bars are set by article 695-22 of the Code de procédure pénale: amnesty extinguishing a prosecution that French courts could have brought, a final decision for the same acts by French courts or by another Member State with sentence executed or no longer enforceable, minority under thirteen at the time of the acts, and, crucially for politically coloured files, discriminatory purpose. The text states: “S’il est établi que ledit mandat d’arrêt a été émis dans le but de poursuivre ou de condamner une personne en raison de son sexe, de sa race, de sa religion, de son origine ethnique, de sa nationalité, de sa langue, de ses opinions politiques ou de son orientation sexuelle ou identité de genre, ou qu’il peut être porté atteinte à la situation de cette personne pour l’une de ces raisons.” Note what this covers and what it does not. A German warrant issued to prosecute a Ukrainian diver because he is Ukrainian, or because of his political opinions, would have to be refused in Paris. A German warrant issued to prosecute the material act of planting explosives under a gas pipeline, whatever the author’s passport or cause, does not trigger this bar. The Polish judge’s reasoning, that the attack was just if attributable to Ukraine, has no equivalent in article 695-22: French law knows no general political-offence exception for the European arrest warrant, only this targeted anti-discrimination shield, and the defence would have to prove that the warrant was issued for a discriminatory purpose, not merely that the prosecution is politically inconvenient. The optional bars of article 695-24 of the Code de procédure pénale then offer six further doors, all discretionary: “Si, pour les faits faisant l’objet du mandat d’arrêt, la personne recherchée fait l’objet de poursuites devant les juridictions françaises ou si celles-ci ont décidé de ne pas engager les poursuites ou d’y mettre fin”, French nationality or residence combined with enforceability of the sentence in France, acts committed in whole or in part on French territory, “Si les faits pour lesquels il a été émis ont été commis, en tout ou en partie, sur le territoire français”, extraterritorial offences that French law would not allow France to prosecute, a final decision for the same acts by a third state, and limitation: “Si les faits pour lesquels le mandat d’arrêt européen a été émis pouvaient être poursuivis et jugés par les juridictions françaises et si la prescription de l’action publique ou de la peine se trouve acquise.” Applied to the diver, these doors look narrow. No French prosecution is pending over the Baltic blasts, the dives did not touch French territory, and the September 2022 facts are nowhere near time-barred. The only theoretically arguable door would be a French decision to prosecute the same facts itself, which would presuppose a French jurisdictional hook that, on current facts, does not exist: the pipelines land in Germany, the boat was rented in Rostock, the explosives were laid near Bornholm, and nothing in the public file points to France.
Dual criminality, the classic extradition battlefield, would not even be fought. Article 695-23 of the Code de procédure pénale states the principle first: “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.” Then comes the exception that swallows this file whole: “un mandat d’arrêt européen est exécuté sans contrôle de la double incrimination des faits reprochés lorsque les agissements considérés sont, aux termes de la loi de l’Etat membre d’émission, punis d’une peine privative de liberté d’une durée égale ou supérieure à trois ans d’emprisonnement ou d’une mesure de sûreté privative de liberté d’une durée similaire et entrent dans l’une des catégories d’infractions prévues par l’article 694-32.” The list in article 694-32 of the Code de procédure pénale runs to thirty-two categories, from participation in a criminal organisation to terrorism, illicit trafficking in arms, munitions and explosives, and, at number thirty-two, a single word that decides the point: “32° Sabotage.” Causing explosions and sabotage of energy infrastructure, punishable in Germany by well over three years of imprisonment, falls squarely into that exempted category, so a Paris chamber would execute the German warrant without any dual-criminality review, the legal characterisation and the applicable penalty being, in the words of article 695-23, a matter for the exclusive assessment of the issuing state’s judicial authority. For completeness, French domestic law would criminalise the same conduct anyway: article 322-6 of the Code pénal provides that “La destruction, la dégradation ou la détérioration d’un bien appartenant à autrui par l’effet d’une substance explosive, d’un incendie ou de tout autre moyen de nature à créer un danger pour les personnes est punie de dix ans d’emprisonnement et de 150 000 euros d’amende.” Ten years: the dual-criminality test would be satisfied even if it applied, which it does not.
Two further lines of defence remain, and both are governed by the demanding 2025 and 2026 case law quoted above. The first is ne bis in idem, the rule against being tried twice for the same acts. The Cour de cassation restated it on 16 July 2026: “Le premier de ces textes prévoit un motif obligatoire de non-exécution du mandat d’arrêt européen lorsque la personne recherchée a fait l’objet, par les autorités judiciaires françaises ou par celles d’un autre État membre que l’État d’émission, d’une décision définitive pour les mêmes faits que ceux faisant l’objet de ce mandat” (Cass. crim., 16 July 2026, no. 26-83.852). And it recalled, through the Paris court’s own reasoning which it endorsed on this point, that the concept of the same acts is autonomous European law: “la notion de « mêmes faits » est une notion autonome du droit de l’Union, et doit, selon la jurisprudence de la Cour de justice de l’Union européenne, 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 (CJUE, GC, arrêt du 16 novembre 2010, Gaetano Mantello, C-261/09, §§ 38 à 41).”, as recalled in Cass. crim., 16 July 2026, no. 26-83.852. For the diver, ne bis in idem currently helps no one: the Polish refusal was a surrender decision, not a final judgment on the merits, and neither France nor any other Member State has finally tried him for the dives. But the July 2026 ruling adds a procedural weapon the defence would certainly use: because the chamber must be able to verify ne bis in idem and the French-proceedings ground, a German warrant that stayed vague on dates, places and material acts would have to be sent back for supplementary information within ten days, and a surrender granted without that verification is annulled, as the partial cassation of the Spanish-file surrender shows. The German file, again, looks robust on this test, but a Paris defence team would still force Berlin to document every dive.
The second line is fundamental rights, and here the August 2025 ruling sets a ceiling that defence teams often misread. The Paris chambre de l’instruction had refused surrender to Romania by invoking, among other things, the age and fragile health of the person concerned, his wife’s fragile health, a child endangered by a long imprisonment, and the disproportionality of the warrant. The Cour de cassation answered on two levels. First, disproportionality review belongs to the issuing state alone, as quoted above. Second, a court that refuses surrender for fundamental-rights reasons must name the specific right at stake and support the risk with objective, precise and circumstantial evidence; it cannot anticipate a Member State’s inability to provide appropriate care in detention, in the words of the fourth ground of appeal, “sans avoir cherché à vérifier l’existence d’éléments objectifs, précis et circonstanciés sur ce point” (Cass. crim., 6 August 2025, no. 25-84.847), and it cannot rely on inoperative considerations such as the age of the facts, health alone, a parallel inheritance dispute or the withdrawal of an Interpol notice, the Court citing the CJEU decisions of 31 January 2023 (C-158/21) and 29 July 2024 (C-318/24) on that exact point. Transposed to the diver, this means a Paris court could not refuse surrender because the war makes the prosecution tragic, because the suspect is a soldier-like figure, or because Ukrainian public opinion views the sabotage as legitimate resistance. It could only refuse on a proven, specific fundamental-rights risk, documented with hard evidence about German detention or trial conditions, or on proof that the warrant was issued for a discriminatory purpose within article 695-22, 5°. That is a steep evidentiary hill, and it explains why, in Paris as in Pula, the likely outcome of this file would be surrender, followed by a trial in Germany under the specialty protection of the warrant. The Polish outcome, however sincerely motivated, would not survive a French pourvoi.
Conclusion
The County Court in Pula has not had the last word: the diver’s lawyer has announced an appeal, and the appellate court will now decide whether the surrender authorisation stands. If it does, Vladimir Z. will join Serhii K. before the German courts, where one suspect denies orchestrating the operation for a state and the other denies diving for it, and where German judges will test whether underwater sabotage of a pipeline landing at Lubmin amounts to the explosions and sabotage charged. For observers in France, the deeper significance lies elsewhere. Within the European arrest warrant system, the same file can produce opposite outcomes depending on the executing court, yet only one of those outcomes respects the instrument: the Polish refusal substituted a judgment on the justice of the cause for the closed list of refusal grounds, while the Croatian authorisation stayed inside it, and the two 2025 and 2026 rulings of the French Cour de cassation confirm that Paris would follow Pula, not Warsaw. A Paris chamber would decide in weeks, not months; it would skip dual criminality because sabotage is on the exempted list; it would demand precise facts and supplementary information within ten days if the warrant were vague; and it would refuse only on a listed ground, a named fundamental right supported by hard evidence, or proven discriminatory purpose. Anyone arrested in France on a foreign European warrant should understand these mechanics before the first hearing, because consent is irrevocable within seven days, specialty can be waived or preserved at that same moment, and the twenty-day contested timetable leaves no room for improvisation. The Nord Stream saga will continue in courtrooms for years; the surrender battle shows that in European extradition practice, procedure is destiny.