On September 29, 2026, the United States Supreme Court allowed the Trump administration to resume flying migrants to countries that are not their own, without giving them a chance to explain what awaits them there. Nine days later, Human Rights Watch published a 100-plus-page investigation documenting what those flights look like: more than 500 people sent to African states between January 2025 and August 2026, shackled for journeys of 18 hours or more, detained on arrival in Eswatini, Ghana and South Sudan, and in several cases passed onward to the very countries whose persecution they had fled. A United States official, quoted by the Senate Foreign Relations Committee, described the purpose of the programme in plain terms: “The point is to scare people.”
For an international audience, the American litigation is only half the story. The other half is the comparison every removal lawyer in Europe is now making: could a French prefect sign the equivalent of an Eswatini flight? The answer, built from the French Code on entry and residence of foreign nationals, three recent rulings of the Conseil d’Etat, two published judgments of the Cour de cassation and the Strasbourg Court’s Article 3 case law, is that Paris operates under constraints Washington does not. France must name a destination in a separate, challengeable decision. It may only choose the country of nationality, a readmission-agreement country, or a country the person agrees to. And no one may be sent anywhere that exposes them to treatment contrary to Article 3 of the European Convention on Human Rights. This article explains the American machine first, then the five French locks that would stop a copy.
The calendar adds pressure. The Senate Foreign Relations Committee has published a staff account of secretive transfer arrangements, and the Supreme Court will hear full argument in December, meaning every flight between now and Christmas operates under a stay rather than a final validation. For practitioners, that interim quality is the most important feature of the American position: the machinery runs while its legality is still being argued, and the people it moves bear the cost of the interval. France, as the second half of this article shows, reverses that presumption by forcing the state to prove the destination lawful before the plane can lawfully leave.
I. How the United States built a third-country removal pipeline
A. Can the Supreme Court let deportations resume without a hearing on the risk?
The September 29 order is an emergency stay, not a final ruling, and that distinction matters. Reuters reported that the Court granted the Justice Department’s request to put on hold the ruling of Boston-based federal judge Brian Murphy, who had held that the Department of Homeland Security policy of removals to so-called third countries is unlawful. The Court, Reuters wrote, let the administration resume deporting migrants to countries other than their own without giving them an opportunity to demonstrate the harm they might face there. The vote followed the Court’s 6-3 conservative-liberal split, with the three liberal justices dissenting, and the justices agreed to hear full arguments in December, with a definitive ruling to follow.
Judge Murphy’s underlying decision had two legs, and both are worth understanding because each has a French equivalent examined below. In February 2026, he ruled that the policy violates immigration-law procedures and the due-process protections of the Fifth Amendment, which generally require notice and an opportunity to be heard before the government takes adverse action. The case reached him as a class action brought by immigrant-rights groups on behalf of people facing transfer to third countries without notice and without a chance to assert the harms they could face there.
The September order was not the Court’s first intervention. In June 2025 it had already suspended an earlier preliminary order by the same judge halting the removals, and in July 2025 it lifted judge-imposed protections for a group of men facing transfer to South Sudan. The administration’s emergency filing called third-country deportations an indispensable instrument for removing certain noncitizens, including dangerous criminal offenders. The destinations already used, according to Reuters, include politically unstable South Sudan, which the State Department tells Americans to avoid because of crime, kidnapping and armed conflict, as well as Uganda, Equatorial Guinea, Liberia and the Central African Republic. The December hearing will decide whether a removal system built on speed can survive a Constitution built on hearings.
The litigation history shows a pattern of interim victories for speed. The June 2025 order suspending the first preliminary halt, then the July 2025 order clearing the way for the South Sudan transfers, each resolved only the question of what happens while the case continues, not who is ultimately right. Judge Murphy’s February 2026 ruling answered the deeper question for the trial court: a system that selects a destination the person has never seen, gives no notice of that selection and offers no hearing on the danger there violates both statutory procedure and the Fifth Amendment’s promise that deprivations follow process. Whether the Supreme Court’s December merits decision preserves that holding or buries it will set the template for every future American transfer deal, including the secretive arrangements now surfacing before the Senate committee.
B. What do the planes, prisons and onward transfers documented in October 2026 show?
Human Rights Watch’s October 8, 2026 report, titled “To Scare People,” converts the procedural dispute into documented human experience. Drawing on ICE deportation data, it states that between January 20, 2025 and August 5, 2026, the United States expelled more than 500 people to African countries that were not their countries of origin. People sent to Eswatini, Ghana and South Sudan were arbitrarily detained there and ill-treated in detention. Cameroon, Equatorial Guinea and Ghana sent some onward to the risk of persecution or torture in their home countries, violating the principle of non-refoulement. In all receiving countries, the report finds, many transferred people were left with no practical option but to return home regardless of the harms awaiting them.
The individual accounts are stark. Tesfay T., a 35-year-old Ethiopian who says he was imprisoned and tortured at home and then tortured again after being kidnapped in Mexico, spent more than a year in American immigration detention. An immigration judge granted him withholding of removal, a bar on returning him to Ethiopia because of the likelihood of torture. In February 2026, while his lawyer prepared a habeas corpus petition, an ICE officer told him to sign a voluntary departure form for removal to Chad. He refused, he said. The officer answered that he did not care and would sign the form himself, and when Tesfay asked to speak to his lawyer, the officer replied that it was too late.
The report describes flights on which people were shackled at wrists, waist and ankles for the entire journey, sometimes 18 hours or more, given little or no information about their destination, denied essential medication and sometimes food and water, with some restrained in straitjackets or a full-body device known as “the wrap.” On arrival, detention followed. Orville Etoria, a Jamaican held at the Matsapha maximum-security prison in Eswatini, described in a deposition for a complaint before the African Commission on Human and Peoples’ Rights weeks without contact with the outside world: no recreation, meals tightly controlled, movement severely limited, detainees made to feel like dangerous criminals and treated as violent offenders although they had done nothing wrong.
Two institutional findings give the report structural weight. First, except for transfers to Cabo Verde, Liberia, Uganda and Rwanda, the removals were not carried out under formal public safe-third-country agreements, and even those four agreements lack the transparency, enforceability, reciprocity and due-process guarantees of the American-Canadian model. Second, the deterrent purpose is avowed: Homeland Security spokesperson Tricia McLaughlin warned in August 2025 that entering the country illegally and breaking its laws could lead to CECOT, Alligator Alcatraz, Guantanamo Bay, South Sudan or another third country, and the unnamed American official told Senate investigators that with countries such as Palau or Eswatini, the administration can threaten people with being “literally dropped in the middle of nowhere.”
The report also documents the incentive architecture around the flights. A carrots-and-sticks programme offers a 3,000-dollar payment, a government-funded one-way ticket and forgiveness of unpaid fines to those who agree to leave voluntarily, a scheme the administration says produced 1.9 million self-deportations by December 2025, while the founding proclamation threatens those who stay with removal, prosecution, incarceration, fines, wage garnishment and confiscation of savings, homes and vehicles. Third-country transfer sits at the punitive end of that menu by design. And the chain does not always stop at the first destination: Cameroon, Equatorial Guinea and Ghana moved some transferees onward toward home-country risks, each link in the chain re-creating the refoulement the first transfer claimed to avoid, in breach of the prohibition that binds every state that lays hands on the person.
II. Could France run the same programme? The five locks in French and European law
A. Why does every French removal need a named, separately challengeable destination?
The first lock is architectural. Under Article L. 611-1 of the Code on the entry and residence of foreign nationals and the right of asylum, the administration may require a foreign national to leave French territory in defined situations, from irregular stay to a final refusal of refugee status. But the removal order alone authorises nothing about the destination. Article L. 721-3 of the same Code provides that “L’autorité administrative fixe, par une décision distincte de la décision d’éloignement, le pays à destination duquel l’étranger peut être renvoyé” whenever an order to leave, a return ban, an expulsion order, a territorial ban or a criminal ban on French territory is enforced. An American-style flight to an unnamed “third country” has no legal existence in this system: without the second decision, there is no lawful destination.
The second lock is the closed list of eligible destinations. Article L. 721-4 allows the administration to designate only three categories: the country of the person’s nationality, another country that has issued a valid travel document under a European or bilateral readmission agreement or arrangement, or, with the person’s agreement, a country where the person is lawfully admissible. A destination picked for its deterrent value, with no readmission instrument and no consent, fails at this threshold. The third lock sits in the final sentence of the same article: “Un étranger ne peut être éloigné à destination d’un pays” where he shows that his life or liberty is threatened there or that he faces treatment contrary to Article 3 of the European Convention on Human Rights of November 4, 1950. Risk, once shown, vetoes the destination.
The fourth lock is judicial, and it was tightened less than a year ago. In its advisory opinion of October 28, 2024, No. 495898, the Conseil d’Etat held that “l’autorité administrative ne saurait légalement désigner comme pays de renvoi” a country where the person risks treatment contrary to Article 3, and that “le moyen tiré de la violation de ces stipulations conventionnelles peut être utilement invoqué” before the administrative courts against the decision fixing the country of return. The same opinion adds the operational duty: “il appartient au préfet” setting the destination to verify, under judicial review, that the measures taken do not expose the person to Article 3 treatment. Litigation since confirms the mechanism works in both directions: on June 5, 2025, the Conseil d’Etat ruled in No. 493675, a case in which the administrative court of Nancy had annulled destination decisions while upholding the underlying removal orders, and on July 12, 2025, its interim-relief judge decided No. 506084, a same-week challenge to two prefectural decisions designating Tunisia as the country of removal. The destination decision can be attacked through the same procedure as the removal order it enforces, which Article L. 721-5 of the Code expressly provides.
The fifth lock is speed itself. Emergency interim relief before the administrative courts can suspend a removal decision within hours when it threatens a fundamental freedom, and the Conseil d’Etat’s July 2025 Tunisia ruling shows that prefects designating a country of return must expect that review. Put together, the French system answers each American shortcut with a checkpoint: a destination must be named, chosen from a closed list, screened for Article 3 risk by the prefect, and defended before a judge who can annul the destination while leaving the removal order intact.
The Vosges litigation behind the June 2025 ruling illustrates how granular the review gets. A father and daughter saw their asylum-application attestations withdrawn, received orders to leave within thirty days, and were assigned a country of return; the Nancy court annulled the destination decisions while leaving the removal orders standing, and the Conseil d’Etat rejected their further appeal. The lesson for both sides is symmetrical: the administration can lose the destination without losing the removal, and the foreign national can win the destination without winning the stay. Because Article L. 721-5 channels challenges to the destination decision through the same procedure as the order it enforces, counsel who attack only the order to leave and ignore the country-of-return decision leave the decisive page of the file unchallenged.
B. What do Strasbourg and the extradition mirror add to the protection?
Above the Code sits the Convention. Article 3 provides, in its authentic English text, that no one shall be subjected to torture or to inhuman or degrading treatment or punishment. The Strasbourg Court has read that short sentence as an absolute bar on sending anyone, however undesirable, to face such treatment. In Soering v. the United Kingdom, it prohibited extradition where there were serious grounds to believe the person would face a real risk of prohibited treatment in the destination country, and in Hirsi Jamaa and Others v. Italy (application No. 27765/09), the Grand Chamber condemned Italy’s pushback of migrants intercepted at sea to a departure country without any examination of individual risk. The American programme’s central feature, transfers without a chance to show the harm faced, is precisely what the Hirsi line of authority forbids in Europe: risk must be assessed individually, before removal, with suspensive effect.
France applies the same logic to extradition, where the Cour de cassation polices foreign assurances with unusual strictness. On January 30, 2024, in No. 23-83.549, the criminal chamber quashed a surrender opinion concerning Tunisia. It held that where the death penalty is at stake, “il appartient à la chambre de l’instruction de s’assurer que l’Etat requérant prend l’engagement dénué de toute ambiguïté que cette peine ne sera pas appliquée à la personne réclamée”: the requesting state must give an undertaking entirely without ambiguity, that the penalty will not be applied to the specific person sought. General considerations, a thirty-year moratorium and renewed international commitments described by the lower court itself as symbolic could not substitute for that individual guarantee, and the Court “CASSE et ANNULE, en toutes ses dispositions, l’arrêt susvisé de la chambre de l’instruction de la cour d’appel d’Aix-en-Provence, en date du 31 mai 2023, et pour qu’il soit à nouveau jugé, conformément à la loi ;” Three years earlier, on January 13, 2021, in No. 20-81.359, the same chamber held that an investigating chamber which finds that the requested person would face a risk of inhuman or degrading treatment if extradited to his country of origin “doit donner un avis défavorable,” and that such a risk is established while the person holds subsidiary protection. A French Eswatini flight would therefore collide twice: administratively, under Articles L. 721-3 and L. 721-4 and the Conseil d’Etat’s 2024 advisory opinion, and criminally, under the Cour de cassation’s demand for person-specific, individually verified guarantees.
The comparison with the American docket sharpens the point. Judge Murphy’s February 2026 ruling rested on notice and hearing; the Conseil d’Etat’s October 2024 advisory opinion rests on the same intuition, allocating the Article 3 objection to the destination decision where it can actually bite. The Supreme Court’s December merits hearing will test whether speed can coexist with due process in American law. In French law that test has already been run, and process won: the prefect must name, justify and defend the destination, the judge can strike the destination down, and Strasbourg watches over both. Readers facing a French removal order with an unfamiliar destination should note the practical consequence the Conseil d’Etat drew: challenge the decision fixing the country of return on Article 3 grounds even when the underlying order to leave looks unassailable, because the two decisions live separate legal lives. A related analysis of how Paris handles incoming American extradition requests, with the same insistence on individualised guarantees, is available in this desk’s examination of the Mohamed-Guyana United States extradition file.
The subsidiary-protection shield in the January 2021 ruling deserves emphasis because it shows French law going further than the bare Convention minimum. The advisory opinion recalls that subsidiary protection goes to a person facing a real risk of death-penalty execution or of torture or inhuman or degrading treatment, decided by the French Office for the Protection of Refugees or the National Court of Asylum and documented by a multi-year residence permit. While that status lasts, the Cour de cassation treats the Article 3 risk as established rather than arguable, closing the file instead of weighing it. No American analogue exists in the third-country context: withholding of removal, the protection Tesfay T. held on paper, did not stop officers from pressing a Chad departure form on him in detention.
Conclusion
The autumn of 2026 has produced a natural experiment in removal law. One system lets the executive choose the destination, move first and litigate later, with the Supreme Court’s December hearing as the only scheduled checkpoint. The other requires the executive to name the destination in a separate decision, pick it from a closed statutory list, screen it for Article 3 risk, and survive a judge who may annul the destination while sparing the removal order. The Human Rights Watch findings on shackled flights, prison detention in Eswatini and onward refoulement from Cameroon, Equatorial Guinea and Ghana show what the first model costs the people it moves. The Conseil d’Etat’s 2024 advisory opinion, its 2025 Tunisia and Vosges rulings, and the Cour de cassation’s unambiguous-assurances doctrine show what the second model demands of the state that moves them. Anyone in France served with an order to leave coupled with an unfamiliar country of return should have both decisions reviewed without delay, because French law treats the destination as a case of its own, and Strasbourg treats the risk there as absolute.