KOHEN AVOCATS FIRM · PARIS 16
Nord Stream Hamburg Trial: Why Poland Refused Extradition While Italy and Croatia Surrendered

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 Hamburg Trial: Why Poland Refused Extradition While Italy and Croatia Surrendered

Speak to a lawyer 06 89 11 34 45

On Wednesday 14 October 2026, a packed public gallery in Hamburg watched the first criminal trial open against a man accused of blowing up the Nord Stream gas pipelines in the Baltic Sea in September 2022. The Hanseatic Higher Regional Court has scheduled two hearings a week until at least February. The prosecution charges the Hamburg defendant, a Ukrainian national in his fifties who denies everything, with a war crime described as an attack by military means against civilian objects, alongside sabotage, causing an explosion and destroying structures. He reached Germany only because Italy surrendered him in November 2025. A second suspect sits in custody in Croatia, where the District Court in Pula approved his surrender to Germany on 24 September 2026 and where he is still fighting it. A third suspect walked free in Poland, where the Regional Court in Warsaw refused surrender on 17 October 2025. Same German European Arrest Warrants, same underlying facts, three opposite judicial answers within one year. For anyone who travels, invests or lives across borders, the split raises a practical question that goes far beyond one pipeline: when an EU country asks for your surrender, what actually decides whether the judge says yes? This article reconstructs the three rulings from court releases and wire reporting, then sets out the French surrender playbook that every requested person meets in Paris, drawn from the Code of Criminal Procedure and the freshest rulings of the Criminal Chamber of the Court of Cassation.

I. Why did Poland refuse to hand over its Nord Stream suspect when Italy and Croatia said yes?

The three cases started from identical German arrest warrants. The Federal Public Prosecutor General in Karlsruhe investigates seven Ukrainian nationals suspected of having chartered a sailing yacht in a German port, sailed it to the Baltic Sea and laid explosive devices on the Nord Stream 1 and 2 pipelines, which were so badly damaged on 26 September 2022 that gas could no longer flow from Russia to Europe. Germany sought each suspect under a European Arrest Warrant for prosecution. What happened next shows how much room the EU surrender system still leaves to the judge of the country where the person is arrested.

A. The Warsaw refusal: no crime under Polish law and no German jurisdiction

On 17 October 2025, the Regional Court in Warsaw refused to surrender the Ukrainian man arrested in Poland and lifted his pre-trial detention. The reasoning, reported in detail by the EU criminal-law review eucrim, rested on two pillars. First, double criminality failed: if the pipelines were attacked as part of a just defensive war waged on behalf of Ukraine, the conduct was not a crime under Polish law, so the condition of punishability in both countries was not met. Second, Germany had no jurisdiction at all in the Polish court’s view: prosecution belonged before an international tribunal for armed conflicts at sea or an ad hoc United Nations tribunal, not before German courts. The Polish judges therefore treated the sabotage as a possible act of war rather than an ordinary crime, and drew the full surrender consequence from that classification.

The Warsaw reasoning matters because it stretches two classic surrender filters to their limit. Double criminality normally asks a modest question: would these facts also be punishable here? Most executing judges answer it abstractly, comparing legal definitions without retrying the case. Warsaw answered it concretely and politically, by importing the law of armed conflict into the comparison. Jurisdiction review goes even further: the executing judge normally trusts the issuing state to delimit its own criminal jurisdiction. Warsaw refused that trust and substituted its own theory of international jurisdiction. Whether other courts follow that path is doubtful, but the ruling is now on the books, and defence lawyers across Europe cite it whenever a warrant touches a conflict zone.

France applies a narrower version of the same double-criminality filter. The Code provides that surrender may be refused where the facts are not an offence under French law, but it removes that control for serious cross-border crime: “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” (Article 695-23 of the Code of Criminal Procedure). The categories named by the Code include terrorism, trafficking in arms, munitions and explosives, organised crime and sabotage-type offences, so pipeline sabotage charged as terrorism or explosives trafficking would pass the French filter without any comparison with French definitions. And the exemption is strict on its face: the Criminal Chamber quashed reasoning that skipped one of its two cumulative conditions, recalling that for offences in the listed categories the court must still confirm the three-year threshold under the issuing state’s law, because “ces conditions étant cumulatives pour l’application de la dérogation au principe de la double incrimination” (Crim., 29 May 2024, No. 24-82.747). The same ruling adds a realistic caveat for requested persons: where no refusal ground is actually raised, “les juges ne sont pas tenus de rechercher d’office si les faits, objet du mandat d’arrêt européen, constituent une infraction au regard de la loi française” (Crim., 29 May 2024, No. 24-82.747). In Paris, the double-criminality argument must be built and filed; the chamber will not build it for you.

One more French lesson flows from the Warsaw impasse. The Polish court freed its suspect by denying German jurisdiction over explosions in international waters. Germany answered that objection on 15 January 2026, when the Federal Court of Justice held that German courts do have territorial jurisdiction because the result of the offence, pipelines rendered unusable, occurred on German territory where the pipelines land. France knows the same effects-based reasoning: an offence partly committed on French soil opens an optional refusal ground rather than an automatic one, since surrender may be refused where the facts were committed in whole or in part on French territory. The lesson for a requested person is that jurisdiction arguments belong to a closed list of refusal grounds, and outside that list they persuade only if the executing court is willing, like Warsaw, to go beyond it.

B. The Italian and Croatian yes: the executing judge does not retry the case

Italy took the opposite road. The Bologna Court of Appeal first authorised surrender of the alleged coordinator of the diving team. On 15 October 2025 the Italian Court of Cassation annulled that order, not on substance but on procedure: Bologna had reclassified the German charges to trigger tighter terrorism procedure, which the supreme court held the executing judge cannot do. After remand, Bologna reordered surrender on 23 October 2025 without revisiting the defence’s deeper objections on immunity, double jeopardy and inhuman-treatment risks, and on 19 November 2025 the Italian Court of Cassation dismissed the second appeal and confirmed surrender. The transfer followed on 27 November 2025. The message is the mirror image of Warsaw: the executing court checks boxes, it does not rehear the file.

Croatia is walking the same path one year later. After German prosecutors announced on 19 August 2026 that a second suspect had been arrested in Croatia on a German warrant, the District Court in Pula approved his surrender to Germany on 24 September 2026. Wire reporting confirms he remains in custody while challenging that approval. Nothing in the Croatian sequence suggests a Warsaw-style review of the merits: arrest on a European warrant, judicial control of the formal conditions, approval of surrender, then the appellate phase. If the challenge fails, Germany will hold two of its seven suspects and try them in Hamburg under the same war-crimes theory.

That war-crimes theory already shapes the Hamburg trial and may yet become its biggest procedural battle. The defence told the court on opening day that the surrender was flawed because the war-crime allegation had not been mentioned when Italy handed the defendant over, and that as a soldier bound to secrecy he cannot properly defend himself or call alibi witnesses. Both points go to the specialty rule, the oldest protective wall in surrender law: a surrendered person answers only for the facts named in the warrant. France writes that wall in plain terms: once surrender is obtained, “celle-ci 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” (Article 695-18 of the Code of Criminal Procedure). The Code then lists only five exits, including the express consent of the executing judicial authority. The Hamburg defence’s specialty argument will therefore be watched in every surrender practice in Europe: if a war-crime label added after surrender can found a conviction, the wall has a hole; if it cannot without fresh Italian consent, the prosecution must either seek that consent or try the defendant on the sabotage and explosives charges as originally surrendered. The French ruling practice on extensions of surrender is strict on exactly this paperwork, as the second half of this article shows.

II. Arrested in France on a European Arrest Warrant: what actually decides surrender?

Suppose the Pula detainee had been arrested in Paris instead of on the Croatian coast. The script would be French, fast and almost entirely judicial. The person appears before the investigating chamber of the Paris Court of Appeal within days of presentation to the prosecutor, since “La personne recherchée comparaît devant elle dans un délai de cinq jours ouvrables à compter de la date de sa présentation au procureur général” (Article 695-29 of the Code of Criminal Procedure). The hearing is public, identity is recorded, statements are taken down, counsel and where needed an interpreter assist, and the issuing state may even be authorised to intervene at the hearing. If the person consents, the chamber records the consent and any waiver of specialty after warning of their irrevocable effects, and where the legal conditions are met it grants surrender within days: “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” (Article 695-31 of the Code of Criminal Procedure). Consent decisions are not appealable. If the person opposes surrender, the chamber rules on the refusal grounds, and its order can go to the Court of Cassation. No minister signs anything: within the Union, surrender is judges judging judges.

A. The defences that work before the French surrender judge, and the ones that fail

The first thing a Paris lawyer tells a requested person is that the refusal grounds form a closed list. The Criminal Chamber repeated it firmly on 16 December 2025: “la remise de la personne recherchée ne peut être refusée que pour les motifs de refus, obligatoires ou facultatifs, qui y sont limitativement énoncés” (Crim., 16 December 2025, No. 25-87.663). Family life in France, however real, is not on that list. A seventy-seven-year-old father of a school-age child, settled in France for years with a fragile state of health on both sides of the household, still saw the Paris chamber’s refusal quashed: the chamber had weighed his private life without finding what the Court demands, because “il lui appartient de rechercher si le demandeur établit l’existence de circonstances exceptionnelles permettant de faire prévaloir le droit au respect de sa vie privée et familiale en France sur le but légitime poursuivi par sa remise en exécution du mandat d’arrêt européen, notamment au regard de la gravité des faits” (Crim., 16 December 2025, No. 25-87.663). Health and family arguments therefore need an exceptional-circumstances file, measured against the gravity of the alleged facts, or they fail.

Double jeopardy works, but only for the same material facts. In a July 2026 ruling on a Spanish terrorism warrant, the Court restated the European definition: same facts means the same concrete circumstances inextricably linked together, whatever the legal labels, since the notion covers “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” (Crim., 16 July 2026, No. 26-83.852). A Paris acquittal for terrorist conspiracy in one place and period did not bar surrender for different apartments, dates and acts, even inside the same terrorist organisation. The same decision protects requested persons in a different way: where the warrant’s information is too thin to rule, the chamber must ask the issuing authority for more, because “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” (Article 695-33 of the Code of Criminal Procedure). Thin warrants can be sent back for homework within a ten-day window for receipt of the answers.

A prior French refusal does not immunise the person either. In July 2026 the Court upheld surrender to Romania on a second warrant for the same sentence after a final 2023 refusal, holding that the earlier order concerned a distinct warrant so that “faute d’objet identique, son autorité ne peut être opposée à l’examen du mandat d’arrêt européen objet de la présente instance”, and adding that “la procédure relative à l’exécution d’un mandat d’arrêt européen émis en vue de l’exécution d’une décision de condamnation ne constitue pas une poursuite nouvelle pour les faits objet de cette condamnation, de sorte que le principe ne bis in idem ne lui est pas applicable” (Crim., 29 July 2026, No. 26-84.072). The practical point is blunt: winning once does not end the file if the issuing state repairs its warrant and comes back.

Two structural features complete the French picture and both cut against Warsaw-style creativity. First, the political-offence exception, the classic shield of extradition law, has no general equivalent inside the European Arrest Warrant system, and French courts apply the listed grounds without inventing a political filter. Second, the chamber does not assess guilt: it checks identity, the warrant’s formal contents, the absence of a listed refusal ground and the punishability thresholds, then surrenders. Anyone expecting the Paris chamber to declare a sabotage lawful, as Warsaw did, misunderstands the job description. The French fight is technical, documentary and fast: attack the warrant’s contents, raise a listed ground with evidence, demand complementary information where the file is thin, and take a lost point to the Court of Cassation on statements of reasons. Readers facing the mirror situation outside the Union, where a state seeks extradition through the classical route rather than a European warrant, find a harsher but slower machine analysed in our guide to how France tests foreign punishments before extraditing, and those tracked through police channels should read our analysis of British warrants and post-Brexit surrender from France.

B. After surrender: extra charges, specialty consent and competing requests

Surrender is not a blank cheque; it is a labelled parcel. The Hamburg specialty row illustrates why. Under French law, trying the surrendered person for earlier facts not named in the warrant needs an extension of surrender, and the French court that ordered the handover decides on it: the investigating chamber “est saisie de toute demande émanant des autorités compétentes de l’Etat membre d’émission en vue de consentir à des poursuites ou à la mise à exécution d’une peine ou d’une mesure de sûreté prononcées pour d’autres infractions que celles ayant motivé la remise et commises antérieurement à celle-ci” (Article 695-46 of the Code of Criminal Procedure). The same article extends the mechanism to onward surrender to another Member State. In April 2026 the Criminal Chamber showed how seriously it takes that paperwork: a man handed over by Spain for one murder case could be investigated for a second, separate murder only after France sought and obtained Spain’s express consent, and the Court approved the extension because the request file recorded his refusal to waive specialty for the new facts. The chamber had noted that he “ne pouvait par conséquent être poursuivi et mis en examen pour les faits objets de la présente procédure, distincts de ceux ayant motivé le mandat d’arrêt européen en exécution duquel il a été remis aux autorités françaises par les autorités espagnoles, qu’avec le consentement exprès de ces dernières” (Crim., 15 April 2026, No. 25-86.561). Applied to Hamburg, the logic is uncomfortable for the prosecution: a war-crime charge materially different from the sabotage and explosives counts named in the Italian surrender documents needs either a fresh Italian consent or a convincing argument that it is the same facts under a new label, and the Mantello test quoted above will decide which of the two it is.

Competing requests add a second layer that the Nord Stream file has so far avoided but that frequent travellers should understand. When a European warrant and a classical extradition request target the same person, different states pull in different directions and someone must choose. EU case law gives the executing authority room to decide which request takes precedence, provided the choice can be reviewed by a court. France itself litigates the classical side of that world every month: red notices, provisional arrest, the race to file papers. In September 2026 the Criminal Chamber quashed an investigating chamber that had declared itself incompetent after a Turkish provisional-arrest file moved courts, recalling that “la mise en liberté de la personne visée par la demande d’arrestation provisoire ne s’oppose pas à une nouvelle arrestation et à son extradition si la demande d’extradition parvient ultérieurement” (Crim., 9 September 2026, No. 25-88.255). Release for want of papers ends that round, not the match: re-arrest elsewhere restarts the case before the newly competent court.

The classical track, for non-EU requesting states, runs slower and more protectively than the European one. Anyone arrested on an extradition request “doit être conduite dans les quarante-huit heures devant le procureur général territorialement compétent” (Article 696-10 of the Code of Criminal Procedure), with counsel, an interpreter where needed and a recorded choice to consent or oppose. Provisional custody has its own guillotine: a person held provisionally “est mise en liberté si, dans un délai de trente jours à dater de son arrestation, lorsque celle-ci aura été opérée à la demande des autorités compétentes de l’Etat requérant, le gouvernement français ne reçoit pas l’un des documents mentionnés à l’article 696-8” (Article 696-24 of the Code of Criminal Procedure). No papers within thirty days of arrest means release, with the file resumable later. The contrast with the European track is the whole story of this article in one line: within the Union, judges move people in weeks on formal checks; outside it, France counts days, demands files and keeps the political decision at the end.

Conclusion

The Hamburg trial will run into 2026 and beyond, the Croatian file is still being fought, and the Polish refusal stands as the dissenting opinion of the European surrender system. Three courts read the same German warrants and produced refusal, surrender and pending review, because the European Arrest Warrant trusts the executing judge with real control and real discretion. France uses that discretion narrowly: listed refusal grounds only, exceptional circumstances for family life, same material facts for double jeopardy, cumulative conditions for the double-criminality exemption, and a specialty wall that needs express consent before any extra charge. Anyone named in a European warrant who is arrested in France should therefore prepare a documentary defence from day one: the warrant’s contents, the exact facts and dates for a double-jeopardy comparison, the evidence for any listed ground, and a specialty strategy that starts at the first hearing, since consent and waiver recorded there are irrevocable. The Nord Stream saga shows what happens when those preparations meet judges who disagree with each other. It also shows, four years after the explosions, that in cross-border criminal law the decisive battle is often fought long before trial, in the surrender courtroom.

📄 Official circular