Erstmals seit 1972 verlassen Astronauten wieder den erdnahen Weltraum. Auf der letzten Etappe ihres Flugs könnte es gefährlich werden.

Über das Fruchtwasser gelangen Aromastoffe zum Ungeborenen. Wie sehr das den späteren Geschmack prägt, ist umstritten.

Allergiker aufgepasst: Sobald die Sonne scheint und die Temperaturen über fünf Grad steigen, beginnt in der Schweiz wieder der Pollenflug. Hier gibt es Antworten auf die wichtigsten Fragen rund um Heuschnupfen, aktuelle Pollenarten und Tipps für den Alltag.

Eine 100-jährige Leistungsschwimmerin, ein 107-Jähriger, der im Supermarkt arbeitet: Wissenschafter versuchen anhand solcher Menschen das Geheimnis eines langen Lebens zu verstehen.

Am Unispital Zürich haben sich zwei Brandverletzte aus Crans-Montana mit einem gefährlichen Bakterium infiziert, einer von ihnen ist gestorben. Experten befürchten weitere Fälle.

“I don’t need international law.” This statement by US President Donald Trump is likely to go down in history, although it remains to be seen whether it will one day serve as evidence of the demise of the international legal order or as testimony to the unprecedented hubris of an authoritarian ruler. In his interview with the New York Times, Trump did not go so far as to deny the binding effect of international law. However, he declared himself the arbiter of when legal boundaries would apply to the US. Only his morals and his mind could stop him.

These statements, made immediately after the United States’ armed intervention in Venezuela, raise questions that go far beyond the classification of the specific operation under international law. The significance of the Venezuela case does not stem from the fact that international law has once again been violated, but from the circumstances surrounding the breach of law: a National Security Strategy published at the end of 2025, which reads like a neo-imperial document; the ongoing deliberations and threats to incorporate territories such as Greenland, and the striking and blunt openness about the economic motives for intervening in Venezuela, without even a pretext of an attempt to justify US actions within the established categories of international law.

The Illegality of the Use of Force

The international law applicable in the case of Venezuela is quickly explained: the US attack, dubbed “Operation Absolute Resolve”, not only disregarded Venezuela’s territorial sovereignty as guaranteed under international law. Above all, the military operation violated one of the cornerstones of the current international legal order, the prohibition of the use of force. Neither did the UN Security Council authorize military action, nor had Venezuela launched an armed attack that would have entitled the US to exercise its right to self-defence. In particular, forms of alleged narco-terrorism, which the US administration pointed to in the context of the various boat strikes, do not come close to triggering the right of self-defence under Article 51 of the UN Charter and customary international law.

It is true that Nicolás Maduro is not a democratically legitimised ruler. However, this fact does not justify the breach of the prohibition of the use of force. It was already highly questionable to deny Maduro’s status under international law as the head of state of Venezuela at the time of the attack, as the US did. This is because Maduro’s position of power remained effective until the US took forcible action. Nor had the international community, for example in the United Nations, agreed to collectively deny Maduro’s status as head of state.

But even if one disagrees with these assessments, this would not alter the illegality of the use of force. In the case of military intervention by invitation, the US would have had to obtain the consent of a person authorised to represent Venezuela before the military operation. There is no evidence of such a request. Whether Maduro was the head of state of Venezuela on 3 January is therefore only relevant to the question of whether the US additionally committed a violation of immunity and whether such a violation has continuing effects in the ongoing criminal proceedings in New York.

Finally, the accusation of systematic human rights violations frequently levelled against Maduro is also irrelevant with respect to the legality of the operation under the framework of the jus ad bellum. According to the prevailing view in international law, this is already the case because the use of military force to end human rights violations requires authorisation by the Security Council. However, even if a so-called humanitarian intervention were considered permissible in an extreme case without such authorisation, this would not justify the use of force in question. According to all available information, the human rights violations in Venezuela did not reach near-genocidal proportions. Moreover, the US attack was declared to have objectives other than ending possible human rights violations – namely, the execution of arrest warrants for drug trafficking, the acquisition of control over Venezuelan oil, and ultimately the consolidation of US hegemony over the continent.

From Unlawfulness to Lawlessness

Any violation of the prohibition of the use of force under international law is serious in and of itself, and this is all the more so if committed by a permanent member of the Security Council. The significance of this particular breach of law, however, has wider ramifications. “Operation Absolute Resolve” testifies to the determination of the US government to descend into lawlessness. Only superficially does this intervention appear to be a mere repetition of the kind of intervention that the US undertook in Panama in 1989/1990. At that time, too, the US intervened militarily in a Latin American state and flew its ruler, General Manuel Noriega, out of the country to prosecute him in the US on charges of drug trafficking. Then, as now, the situation under international law was relatively clear. Yet the US did put forward legal arguments that were at least debatable. Panamanian security forces had attacked US soldiers in Panama, and General Noriega had declared war on the US in the run-up to the intervention. The first Bush administration sought to present its actions as compliant with international law and to put forward the most plausible argument possible under the circumstances. Even in the case of the 2003 Iraq War, whose illegality had triggered considerable criticism worldwide in academic and public debate, the US presented a justification that was not entirely untenable under international law.

By contrast, there was no sign of any effort to argue in line with established criteria of international law when Trump, Hegseth and Rubio appeared before the press the morning after the conclusion of “Operation Absolute Resolve”. None of the three mentioned the words “international law”. Instead, the undisguised message was that it was all about oil. It is difficult to recall a moment since the UN Charter came into force when the government of a member state has so blatantly failed to justify its violation of the prohibition of the use of force as the United States did on 3 January this year.

Anyone who had hoped that the US would show a different face when the UN Security Council convened for an emergency meeting two days later was disappointed after the US representative took the floor. While he refrained from explicitly describing ‘Operation Absolute Resolve’ as an armed operation to seize Venezuelan oil, his description of the military attack as a “surgical operation to execute two US arrest warrants” offered no cause for relief. This was not primarily because of the echoes of Putin’s cynical cloaking of Russia’s war of aggression as a “special military operation”. Above all, the fact that the US military operation was based solely on a claim of enforcing US criminal justice confirmed the disregard for international law. US law was elevated to the sole decisive standard worldwide.

A New Vision of Order

The case of Venezuela fits into larger trends that distinguish it significantly from previous US military interventions and highlight the profound upheavals we are facing: This intervention is not about interest-driven realpolitik that overstretches and abuses legal arguments; it is not a matter of shaping the law in the interests of the hegemon. Rather, one might be tempted, along with Henry Kissinger, to see Trump as a “historical figure” who is putting the final nail in the coffin of an era that has come to an end, without offering an alternative model of order to replace it. In fact, the international (legal) order has been subject to profound criticism for some time now due to systemic legitimacy deficits perceived by a variety of actors. These concerns range from international law’s historical involvement in projects of colonialism and imperialism to current debates about double standards in the enforcement of international law. In addition, the limited capacity for reform of the international system contributes to the impression held by some that international law is basically operating on the basis of outdated criteria that lack representativeness, particularly with regard to institutions such as the UN Security Council. In this light, Trump’s foreign policy, as Daniel Marwecki recently pointed out, could act as a catalyst, breaking down traditional certainties and entrenched path dependencies in an unorthodox way and clearing the way for something new.

However, while Kissinger’s thesis of a shift towards the uncertain could claim a certain plausibility under the first Trump administration, the measures taken by the second Trump administration do indeed point to a new vision of order. This vision is characterised by a departure from the principles of political rule as they have developed since early modernity and gives rise to fears of a dissolution of any power limits for a few states or even individuals. The emerging understanding of political rule at the international level combines feudalistic practices and mafia-like methods with ideas of private corporate governance, which unfold against the backdrop of a libertarian understanding of an economy that is based on barely concealed cliques. In this context, the spectres of imperialism and colonialism reappear. As a result, the idea of statehood as legally founded and limited, and as committed to the common good, is ultimately renounced; ideas that are reflected at the level of international law: the legitimisation of the exercise of power in the pursuit of common public interests, the limitation of sovereignty through the allocation of responsibilities and the demarcation of competences, the recognition of other actors and their interests as legitimate, the differentiation between public and private, between internal and external, between war and peace. The latter distinctions in particular have been under considerable pressure for some time now for a variety of reasons, but the Trump administration’s foreign policy actions seem to be seeking to permanently undermine any understanding of political rule that also aims to contain power.

Signs of this include not only the considerable personal threats against Venezuela’s interim president. The statements made to Denmark regarding Greenland, particularly in the run-up to the World Economic Forum in Davos, as well as the dismissive treatment of Ukrainian President Zelensky in the White House, also reflect the explicit hierarchisation of the global state system reminiscent of feudalism, in which other states are dependent on Trump’s personal goodwill. The fundamental distinction between public authority and personal property is dissolved when the American president attempts to derive private benefit from his public duties. This applies to his efforts to obtain control over Greenland (or at least parts of it) not only for security reasons, but also with the aim of exploiting its resources. Moreover, the construction of the peace plan for Gaza shows how much the exercise of public authority is mixed with private action. Unlike comparable cases in the past, the role of international institutions, such as the World Bank, is being diminished. At the same time, a “Board of Peace” headed by Donald Trump is to manage a “special economic zone” based on ideas of private corporate governance, thus creating spaces that are gradually slipping out of public control.

A Pre-Legal Understanding of Sovereignty

What is more, the Board of Peace seems to have morphed into a more general vehicle for realizing the Trump Administration’s new vision of world order. The foundational Charter of the Board does not even mention Gaza. Instead, it is now styled as a self-standing international organization with a highly unusual conception of governance. This includes the twofold role of Donald Trump as Chairman in a personal capacity and possibly beyond the end of this term in office as US President, as well as his role as the representative of the United States on the Board. The decisions of the Board are subject to his approval and he “is the final authority regarding the meaning, interpretation and application of this Charter” (Art 7). While it remains to be seen how impactful this dystopian vision of a highly privatized international organization will become, the damage is already done: a permanent member of the UN Security Council is attempting to launch a new – almost autocratic – form of multilateral organization in outright competition to the UN system as such.

These developments come together with a pre-legal understanding of sovereignty that exempts the US president from (international) legal obligations by elevating him, in the sense of Carl Schmitt, to the role of an unbound guardian of his nation’s welfare. Numerous statements by President Trump, Vice President Vance, or White House Deputy Chief of Staff Miller reflect this understanding, for example when the US president says, “He who saves his country does not violate the law.” The President expressed a similar view in the New York Times interview referred to in the introduction. Observers describe this understanding of power in various combinations as feudal, mafia-like, imperialistic or even fascist. Ultimately, these historical comparisons and metaphorical labels serve to make tangible a paradigm shift that is unfolding before our eyes. The fact that political language has not yet developed precise terminology to describe this foundational change does not justify considering the current situation as “business as usual” in light of past US intervention practices.

Europe’s Choice

What does this mean for German foreign policy and that of other European states? In German foreign policy discourse, some voices argue that Europe can no longer afford to stand up for international law in view of its military and digital weakness in the new global political constellation. According to this view, the decision to keep the US on board in defending Ukraine must be the deciding factor. In light of the concurrence of a devastating Russian war of aggression and the rejection of legal obligations by the US in international politics, it is supposedly a matter of political wisdom to put international law on the back burner for the time being in order to avoid open contradiction with the Western alliance’s foremost power. From this perspective, the German Chancellor’s incorrect classification of the Venezuela mission as “legally complex” was understandable in light of the “politically complex” situation for Europe.

Those who tend to share this view should bear in mind that, with its statements on Greenland, the US administration has already set its sights on European territory. While the Davos conference may have provided some form of détente for the time being, it remains to be seen how the announced agreement on greater US influence over parts of Greenland will unfold and whether the outcome of these negotiations will placate the US administration. A glance at the US National Security Strategy of December 2025 shows that Greenland may in any case not be the end of it. In that document, Europe is treated – cynically put – no differently than the rest of the world. The security strategy announces, in a barely veiled manner, that the new MAGA establishment in the US can also envisage “regime change” in Europe in order to halt its supposed civilisational decline, which European governments are said to foster, inter alia, by allegedly censoring freedom of expression.

After the case of Venezuela, there can be no doubt that the announcements in the National Security Strategy are more than just rhetoric. Thus, with every violation of international law by the US on other continents, which Germany and other European states do not call out for fear of a reaction from the US President, the danger increases that the US’s behaviour will have further repercussions. Above all, powers with imperial tendencies that, unlike Russia, have not yet permanently abandoned the path of compliance with international law may feel challenged to emulate the United States. At the same time, German and European silence on US violations of international law threatens to reduce the group of states whose solidarity European states can count on should the US or other imperial powers turn against them. States such as Brazil, Mexico, Colombia and South Africa are likely to have felt abandoned by Europe during the Security Council debate on the Venezuela case. The fact that seven European heads of government immediately thereafter emphatically invoked international law with regard to Greenland did not improve matters regarding the “Global South”. If Europe applies double standards depending on whether the case concerns Venezuela or Greenland, disappointment or resentment in the “Global South” is inevitable. It is not in Europe’s interest to provoke such reactions time and again. Europe will need a coalition, particularly with democratic states from the ‘Global South’, to counter the neo-imperial logic of US and Russian actions, with the possibility of China joining them. This should include a willingness to make compromises with the “Global South” on contested international legal policy issues, for example when it comes to the role of the UN General Assembly in the UN system or policy decisions on the future direction of international economic law.

Eventually, Germany and the other European states cannot afford to abandon international law – and the international legal order cannot be preserved with only selective commitment to it. Germany should therefore join forces with as many other states as possible to signal to the US that it is not prepared to go down the path of lawlessness in international relations, regardless of which continent is involved. Obviously, this is a highly inconvenient stand to adopt in view of Germany’s own relative military and digital weakness. But if countries such as Colombia and Mexico, which are directly exposed to the threats of the US president, are prepared to uphold international law in the Security Council, then Germany, France, and the United Kingdom should also be prepared to do so.

What Can Europe Offer?

In this undertaking, European countries should remember their strengths. Even under the new global political circumstances, Europe remains a player whose economic weight cannot be ignored for the time being. More importantly, Europe stands for a normative alternative that holds global appeal in contrast to the imperial-feudalistic US vision of order. The attractiveness of the set of ideas to which Europe is indebted is closely linked to the call for legally justifying the exercise of political power and to the legal limitations placed upon it, both internally and externally. Europe should advance its normative offer at the international level, especially vis-à-vis states of the “Global South,” and forge new coalitions with them in the interests of freedom, equality and solidarity. In that context, it is a hopeful sign that trade relations appear to be in the process of being strengthened between the European Union and Latin American States (Mercosur) as well as India.

On 18 January, eight European states declared regarding the escalation of the Greenland conflict: “We are committed to upholding our sovereignty.” That is precisely the point. Canadian Prime Minister Mark Carney clearly stated the alternative at the World Economic Forum in Davos on 21 January: “This is not sovereignty. It’s the performance of sovereignty while accepting subordination.” This would be the consequence if middle powers stopped defending international law. The sovereignty of militarily weaker states ultimately arises only from international law. To defend the international legal order is thus in our own best interest.

The initial version of this article appeared in German in the Frankfurter Allgemeine Zeitung on 22 January 2026 (p. 6); the English version has been updated to reflect developments until 30 January 2026. The translation was prepared with the help of DeepL.

The post The End of an Era? appeared first on Verfassungsblog.

Die US-amerikanische Immigration and Customs Enforcement Agency (ICE) hat angekündigt, die „Ad-Tech- und Big-Data-Tools“ privater Unternehmen für die Verfolgung vermeintlich illegaler Einwanderer nutzen zu wollen. Die Forderung zeigt in schonungsloser Offenheit, wie groß das Missbrauchspotenzial von Geschäftsmodellen ist, die auf massenhafter aggressiver Datenextraktion beruhen. Gepaart mit der ökonomischen Struktur des heutigen Internets und der Realität des Technologie-Oligopols in der Hand weniger Firmen entfaltet sich ein dystopisches Potenzial: Private Datenmacht verbindet sich mit einer Behörde, die über erhebliche Vollzugs- und Zwangsbefugnisse verfügt. Darin zeigt sich eine neue Stufe der Erosion demokratischer Kontrolle und Rechtsstaatlichkeit.

Datenmacht und Vorhersagetechnologien als Risiko

„AdTech“ (Advertising Technology) bezeichnet das Geschäftsfeld der Big-Data- und KI-Industrie, in dem digitale Werbung automatisiert ausgespielt, gemessen und optimiert wird. Besser bekannt als „personalisierte Werbung“ handelt es sich dabei um eine seit mehr als 20 Jahren gewachsene Industrie, die auf der systematischen und flächendeckenden Sammlung von Tracking- und Verhaltensdaten im Internet beruht. Anhand der Datenspur einer Nutzer:in werden mit statistischen Verfahren Persönlichkeitsmerkmale ermittelt – Bildungsstand, Interessen, psychologische Profile, Krankheitsrisiken, demographische oder sozioökonomische Informationen – anhand derer dann entschieden wird, welche Werbung für diese Nutzer:in am „effizientesten“ ausgespielt werden kann.

Solche Modelle lassen sich prinzipiell auf jede beliebige Person zur Vorhersage von Informationen anwenden, die diese Person nie irgendwo preisgegeben hat. Dies prägt die erhebliche informationelle Machtakkumulation, die mit AdTech verbunden ist. Seit den 2000er-Jahren ist die Industrie bemüht, das Image zu verbreiten, dass personalisierte Werbung gesellschaftlich harmlos sei. Der potenzielle Zugriff einer Behörde wie ICE auf diese technologischen Machtinstrumente zeigt jedoch das verheerende Missbrauchspotenzial dieser Technologie.

Dieses Missbrauchspotenzial realisiert sich bei der Zweckentfremdung ursprünglich kommerzieller und ziviler Technologie für den behördlichen Vollzug. Darin liegt eine gefährliche Vermischung des staatlichen Gewaltmonopols mit weitgehend unregulierter privater Wirtschaftsmacht. Bei den statistischen Verfahren der AdTech-Industrie handelt es sich um prädiktive Machine-Learning-Systeme, die massenweise Vergleichsdaten benötigen, um trainiert zu werden. Deshalb liegt der AdTech-Markt fest in der Hand weniger großer Plattformunternehmen, die über Milliarden Nutzer:innen und somit über die Möglichkeit der Datenaggregation verfügen – darunter prominent Google, Meta, Apple, Amazon und Microsoft.

Wer also AdTech betreibt, der hat eine ganz besondere Form von Datenmacht, die wir Vorhersagemacht nennen. Diese besteht aus informationstechnischen Apparaten, die die Vorhersage nahezu beliebiger Persönlichkeitseigenschaften von Individuen anhand von vergleichsweise leicht erhältlichen Nutzungsdaten (Tracking, Social Media, Apps) erlauben. Der Katalog der Datenpunkte, die große Plattformen routiniert über alle ihre Nutzer:innen abschätzen, um sie für personalisierte Werbung zu verwenden, umfasst hunderte geschätzter Attribute, von religiösen Neigungen und ethnischer Herkunft bis zu Substanzmissbrauch, Beziehungsstatus oder politischen Ansichten. Der EuGH hat dazu ausgeführt, dass die Datenverarbeitung von Meta erhebliche Auswirkungen auf Nutzer:innen hat, deren Online-Aktivitäten zum großen Teil, wenn nicht sogar fast vollständig, von Meta aufgezeichnet werden, was bei Nutzer:innen das Gefühl auslösen kann, dass ihr Privatleben kontinuierlich überwacht wird.

ICE, Daten und KI-Technologie

ICE ist eine finanziell überproportional gut ausgestattete Behörde, die dieses Budget auch bisher dazu genutzt hat, unterschiedliche Digitaltechnologien zu erwerben. Wäre ICE eine Armee, läge es auf Platz 14 der bestfinanziertesten Militärs der Welt. Bereits seit einigen Jahren setzt ICE auch auf zunehmend digitale Überwachungstechnologien im alltäglichen Einsatz; die Behörde hat verschiedene Verträge zur Überwachung von Aufenthaltsorten, Gesichtserkennung, Spyware, Social-Media- und Telefonüberwachung abgeschlossen. Um zwei Beispiele aufzugreifen: ICE nutzt Software, die Standortdaten von Millionen von Menschen von Datenhändlern sammelt, um diese mit öffentlichen Social-Media-Daten zu verbinden. Die Firma Penlink verkauft diese Informationen dann an ICE, damit die Behörde ohne eine gerichtliche Entscheidung Kenntnisse über Aufenthaltsorte und Standortdaten erlangen kann. Ein weiteres von ICE genutztes KI-Tool nennt sich Onyx und wurde von der Firma Fivecast entwickelt. Onyx wird beworben als „KI-gestützte Risikoanalyse – alles auf einer Plattform. Sie erhalten Zugriff auf Milliarden von Datenpunkten aus Millionen von Quellen, sodass Sie Bedrohungen schneller und genauer aufdecken können.“

Wie in allen anderen Einsatzbereichen auch, sind diese Technologien fehleranfällig. Die von ICE verwendete Gesichtserkennungs-App hat mehrfach Personen falsch identifiziert – Kontrollmechanismen oder Qualitätssicherungsvorgaben sind nicht bekannt, die Grundrechtsrelevanz im Kontext der Entscheidung über eine Abschiebung im Schnellverfahren ist offensichtlich.

Machtmissbrauch durch Zusammenarbeit mit kommerziellen Diensten

Im Zuge der technologischen Hochrüstung von ICE markiert die aktuelle Anfrage an Unternehmen der US-amerikanischen AdTech-Industrie eine neue Eskalationsstufe. Erstmals zielt die Behörde explizit darauf ab, Datenbestände und Vorhersagetechnologien zu nutzen, die für kommerzielle und zivile Zwecke entwickelt wurden. Konkret sollen Modelle zur Vorhersage von Persönlichkeitsmerkmalen, die bislang der Personalisierung von Werbung dienten, zur Identifikation und Verfolgung vermeintlich illegaler Migrant:innen eingesetzt werden.

Da es sich bei den KI-Systemen der AdTech-Industrie um besonders leistungsfähige, mit enormen Datenmengen trainierte Modelle handelt, stellt ihre staatliche Nutzung eine schwerwiegende Bedrohung für Freiheit und Grundrechte dar. Hier verbindet sich eine weitgehend unregulierte private Daten- und Vorhersagemacht mit hoheitlichen Zwangsbefugnissen. Die Zweckentfremdung ziviler Technologien für behördlichen Vollzug offenbart einen grundlegenden Machtmissbrauch und den nächsten Schritt der Erosion rechtsstaatlicher Garantien.

Ein warnendes Beispiel für Europa

Über den konkreten US-amerikanischen Kontext hinaus ist dieser Fall insbesondere aus europäischer Perspektive von Bedeutung. Entscheidend ist nicht allein das Handeln von ICE, sondern das gesellschaftliche und systemische Risiko, das bereits in der bloßen Existenz der zivilen technologischen Infrastruktur der AdTech-Industrie angelegt ist. Die dort stattfindende Akkumulation informationeller Vorhersagemacht birgt inhärent die Gefahr missbräuchlicher Sekundärnutzung und eines schleichenden „function creep“.

Um Gesellschaften wirksam vor diesem Risiko zu schützen, muss dieses in einer präventiven Herangehensweise als systemisches Risiko analysiert werden. Ohne regulatorische und organisatorische Maßnahmen, die Daten und KI-Modelle dauerhaft an ihren ursprünglichen Verwendungskontext – hier die Personalisierung von Werbung – binden, können solche mächtigen Werkzeuge in veränderten politischen Konstellationen jederzeit für grundrechtsgefährdende Zwecke eingesetzt werden. Genau in diesem Missbrauchspotenzial, das überall dort entsteht, wo der Einsatz mächtiger Technologien nicht hinreichend begrenzt ist, liegt eine Dimension systemischen Risikos, die in der europäischen Politik bislang weitgehend verkannt wird. Der Fall ICE ist daher nicht nur ein US-amerikanisches Problem, sondern ein warnendes Beispiel dafür, welche Gefahren auch europäischen Rechtsstaaten drohen, wenn Zweckbindung und demokratische Kontrolle datenbasierter Geschäftsmodelle nicht wirksam abgesichert werden.

Zweckbindung als Grundrechtsschutz

Das rechtliche Instrument der Zweckbindung zielt darauf ab, genau solche Missbrauchsgefahren von Datenmacht einzudämmen. Zweckbindung als Prinzip kann in unterschiedlichen rechtlichen Kontexten verankert sein; es findet sich im US-amerikanischen Kontext bereits in den Fair Information Practice Principles (FIPPS) von 1973. Die USA haben allerdings weiterhin kein Datenschutzgesetz auf Bundesebene, das solche Sicherungsmechanismen verbindlich verankert.

Im europäischen Recht ist Zweckbindung ein Grundpfeiler des Datenschutzrechts und entscheidende Voraussetzung dafür, eine sekundäre Nutzung von Daten nach dem Vorgang ihrer ersten Erhebung, Ableitung oder sonstigen Erzeugung überhaupt kontrollieren zu können. Trotz expliziter Normierung in der DSGVO entfaltet der Zweckbindungsgrundsatz in der Realität nicht seine intendierte Schutzwirkung.

Der vorliegende Fall der AdTech-Industrie illustriert jedoch ein in der Literatur bereits viel diskutiertes Problem mit der Zweckbindung im Kontext von KI: KI-Modelle zur Vorhersage von Persönlichkeitseigenschaften können mit anonymisierten Nutzungsdaten trainiert werden und entziehen sich dann dem Schutzregime des Datenschutzes. Auch die trainierten Modelle selbst, als abgeleitete Datensätze, sind als anonymisierte und abgeleitete Daten regelmäßig nicht unter das Zweckbindungsgebot zu fassen. Deshalb ist der Zweckbindungsgrundsatz im Kontext von prädiktiver Analytik und KI praktisch entwertet worden: Er ist angesichts ubiquitärer Massendaten, zahlreicher beteiligter Akteure und unüberschaubarer Datenverarbeitungsvorgänge inklusive Anonymisierung nicht mehr anwendbar.

Eine Ausweitung der Zweckbindung auf KI-Modelle wäre eine Voraussetzung, um betroffenen Personen und Aufsichtsbehörden überhaupt eine Kontrolle über Datenverarbeitungsvorgänge zu ermöglichen. Denn Daten werden heute nur selten einmal erhoben und ausschließlich für einen bei der Erhebung festgelegten Zweck verarbeitet. In der Industrie ist vielmehr das Gegenteil der Fall: Daten und die aus ihnen trainierten KI-Modelle werden sekundär weiterverarbeitet und in anderen Bereichen monetarisiert. Function creep ist Teil des Geschäftsmodells.

Wenn nun, wie am Beispiel von ICE deutlich wird, staatliche Stellen an den KI-Modellen interessiert sind, die aus alltäglichen zivilen Datenschätzen trainiert werden, erschließt sich damit für die Industrie eine neue, vergleichsweise stabile und lukrative Einnahmequelle.

Grundlage von Datenschutz als Autonomie- und Privatheitsschutz sind zwei zentrale Elemente: zum einen die Möglichkeit, Kontrolle über die eigenen Daten auszuüben, und zum anderen der Grundsatz der „reasonable expectation of privacy“, also angemessene Erwartungen an die Privatsphäre. Beide Elemente bleiben in dieser Konstellation nicht gewahrt. Das stellt bereits im Bereich der zivilen Datenverarbeitung ein großes Problem dar, kann jedoch durch die Sekundärverwendung durch öffentliche Stellen wie ICE besonders gravierende Schäden verursachen.

Die von ICE verfolgten Personengruppen müssen um ihre körperliche Unversehrtheit und ihr Leben bangen – ebenso wie inzwischen Demonstrationsteilnehmende. Dies schlägt zugleich die Brücke zur Demokratie: um freie Diskussionen, Äußerungen, Versammlungen und letztlich Wahlen zu ermöglichen, müssen Bürger:innen in Rechtsstaaten bisher nicht damit rechnen, dass jegliche in völlig anderen Kontexten getätigte Äußerung, jahrzehntealte Facebook-Profile oder private Chats in behördlichen Verfahren herangezogen werden, zu denen sie keinerlei inhaltliche oder kontextuale Verbindung haben.

Zweckbindung operationalisierbar machen

Wir haben deshalb an unterschiedlichen Stellen dafür argumentiert, die Zweckbindung außerhalb des Datenschutzrechts, das stets einen identifizierbaren Datenverarbeitungsvorgang voraussetzt, zu verankern. KI-Systeme selbst, nicht individuelle Datenpunkte und ihr Personenbezug, müssen als Ansatzpunkt der Zweckbeschränkung herangezogen werden. Konkret bedeutet das: Die missbräuchliche Sekundärnutzung von Modellen der AdTech-Industrie muss verhindert werden, indem die Vorhersagemodelle selbst auf ihren zivilen Verwendungskontext festgeschrieben werden. Eine Zweckbindung allein in Bezug auf die Verarbeitung der Daten, die zur Entwicklung dieser Modelle verwendet wurden, ist demgegenüber nicht praktikabel.

Eine überprüfbare und damit vollziehbare Zweckbindung für KI-Modelle, insbesondere im Kontext von Hochrisikosystemen oder -akteuren, würde einen wichtigen Baustein eines effektiven Grundrechtsschutzes liefern. Zwecke müssen transparent kommuniziert, dokumentiert und nachvollziehbar überprüft werden. Ein erster Schritt, dies in der aktuellen EU-Digitalgesetzgebung zu verorten, wäre die Aufnahme von Zwecken und ihre Änderungen im Rahmen der KI-Datenbank nach Art. 71 KI-VO.

Darüber hinaus halten wir es für essentiell, bestimmte „Koppelungsverbote“ zu normieren. Werbe-IDs ermöglichen etwa die Erstellung umfangreicher Profile über Krankheitsprädispositionen, Sucht, sexuelle Orientierung, politische Einstellung, Konsumpräferenz und zahlreiche andere Faktoren, die in der polizeilichen Ermittlungsarbeit schon aufgrund von Diskriminierungsverboten keine Rolle spielen und zudem auch keine kriminologische Relevanz entfalten dürfen. Eine undifferenzierte Nutzung webseitenübergreifender Profile, wie z.B. durch die Meta Business Tools, sollte der Polizei nicht erlaubt sein.

Der Zweckbindungsgrundsatz sollte daher auch in der öffentlichen Beschaffung und bei der Schaffung neuer Kompetenzen stärkere Berücksichtigung finden. Staatliche Stellen sollten keine Systeme nutzen, deren Datengrundlage und vorherige Zweckbindung unklar ist oder die aus vollkommen anderen Anwendungskontexten stammen.

Datennutzung durch die deutsche Polizei

Auch in Deutschland ist der Ausbau polizeilicher Datennutzungsbefugnisse ein politisches Dauerthema. Ohne diese Diskussion in die Nähe der Vollzugspraktiken von ICE zu stellen, liegt in staatlicher Datenmacht ein Missbrauchsrisiko. Das zeigen bereits Fälle, die noch vergleichsweise analog anmuten, etwa die Abfrage von Handynummern zu privaten Zwecken. Solche Risiken potenzieren sich durch den Einsatz von KI-Technologien und durch die Zusammenarbeit mit Unternehmen wie Palantir.

Aktuelle Gesetzesänderungen in mehreren Bundesländern sehen einen erheblichen Ausbau der Datenverarbeitungsbefugnisse der Polizei vor, u.a. auch zum Training von eigenen KI-Systemen. So erlaubt etwa der neue § 24b PolG NRW eine Weiterverarbeitung von Daten zu IT-Entwicklungszwecken: Vorhandene sowie allgemein zugängliche Daten können zur Entwicklung, Überprüfung, Änderung oder zum Training von IT-Systemen verwendet werden. Darin liegt eine erhebliche Aufweichung des Zweckbindungsgrundsatzes, denn bisher war er an vergleichbare Straftaten oder den Schutz ähnlicher Rechtsgüter angeknüpft, die der ursprünglichen Erhebung zugrunde lagen.

§ 42d ASOG-G Berlin ist ähnlich angelegt, geht jedoch deutlich weiter. Die Vorschrift soll Polizei und Feuerwehr ermöglichen, Daten zum Testen und Trainieren von KI-Systemen zu verwenden, und zwar auch über die vorgesehene Speicherdauer hinaus. Absatz 3 sieht die Weitergabe von Daten an Auftragsverarbeiter und sogar an Dritte vor. Die Vorschrift höhlt damit die Zweckbindung aus, indem sie den unbestimmten Zweck „KI-Training“ ohne hinreichende rechtliche Konkretisierung legitimiert. Mangels Differenzierung nach Eingriffsintensität, Kontext oder Sensibilität der Daten sowie aufgrund des dehnbaren Vorbehalts der technischen Machbarkeit drohen zentrale Schutzmechanismen wirkungslos zu werden.

Ausblick

Der Druck auf staatliche Stellen, privatwirtschaftliche Überwachungs- und Vorhersagetechnologien zu nutzen, wird in der Zukunft eher zu- als abnehmen. Der Fall ICE zeigt, dass die Sekundärnutzung datengetriebener, ziviler Systeme – auch in rechtsstaatswidrigen Verfahrensweisen – keine hypothetische Gefahr, sondern eine ernstzunehmende Bedrohung in modernen Demokratien darstellt.

Um einerseits bei der alltäglichen Nutzung digitaler Medien die Voraussetzungen freier Meinungsbildung zu bewahren und zugleich das Potenzial für erhebliche Grundrechtseingriffe durch Missbrauch dieser mächtigen Technologien präventiv einzudämmen, sind neue und weitreichendere Schutzkonzepte erforderlich. Der gegenwärtige Stand der Datenschutzgesetzgebung und KI-Regulierung ist hierfür nicht ausreichend.

Vorsorgendes Handeln ist auch deshalb geboten, weil extensive Ausprägungen von Datenmacht nur schwer zurückgenommen werden können. Technologien, die zunächst durch vertrauenswürdig erscheinende private Akteure und für ein legitimes Ziel eingesetzt werden, können bei wechselnden politischen Mehrheiten schnell missbraucht werden. Die latente Möglichkeit, dass zivile Datentechnologie für hoheitliche Aufgaben missbraucht wird, führt zu einem gesellschaftlichen chilling effect auf freie Entscheidungen, Diskussionskultur, Wahlen und Grundrechtsausübung. Diese Abschreckungseffekte auf Freiheitsausübung beeinträchtigen das Gemeinwohl. So sieht es auch das Bundesverfassungsgericht: „weil Selbstbestimmung eine elementare Funktionsbedingung eines auf Handlungs- und Mitwirkungsfähigkeit seiner Bürger gegründeten freiheitlichen demokratischen Gemeinwesens ist.“

Beide Autor:innen haben im gleichen Umfang zu diesem Artikel beigetragen.

 

The post Von personalisierter Werbung zur staatlichen Verfolgung appeared first on Verfassungsblog.

The General Court (GC) dismissed the action brought by the Association of Jurists for Respect for International Law (JURDI) under Article 265 TFEU in September 2025, amid the humanitarian crisis in Gaza resulting from the Israeli offense. JURDI requested the Court to declare that the EU institutions had failed to act in response to the situation in Gaza in accordance with the positive obligation to prevent genocide. In particular, JURDI argued, the EU should have suspended the Euro-Mediterranean Agreement establishing an association between the EU and Israel, triggered appropriate sanctions, audited EU-Israel cooperation programmes, and issued a clear public position. Admittedly, the Court’s decision to deny locus standi to JURDI is consistent with the case law concerning individuals’ access to court under Articles 263 and 265 TFEU, notable for its strictness. However, aware of the limitations set out in the Treaties, we believe that there is a space for a more liberal interpretation of the rules of access to the CJEU, rooted in the core values of the EU. We argue, therefore, that it is high time the Court reconsider its stance on locus standi and move towards an inclusive and participatory model of legality review that would allow for actions to be brought by non-governmental organisations in the public interest of human rights compliance.

“Europe needs to do more.” Does this also apply to the CJEU?

On 5 September 2025, during the State of the Union, Ursula von der Leyen, president of the European Commission, stated that “what is happening in Gaza is unacceptable” and that in response to European citizens’ concerns “Europe needs to do more”. Thus, she declared that the Commission would propose “a package of measures to carve out a way forward”.

In reference to the statement by the President of the Commission, one might ask whether the CJEU could and should have done more in response to concerns voiced by citizens in the form of action for failure to act. The decision in the JURDI case is, however, yet another chapter in the ongoing saga of the CJEU’s intransigence regarding the standing of private applicants to lodge actions based on Articles 263 and 265 TFEU. Although the GC’s decision should come as no surprise, it is reasonable to question it in light of the core values of the Union, such as human rights and the rule of law, and the obligation to uphold these values in the external actions, as stated in Articles 3(5) and 21 TEU.

Background of the case

The ongoing armed conflict in the Gaza Strip has given rise to questions regarding Israel’s respect for international human rights in view of the unprecedented scale of intervention and its impact on the civilian population. In its Advisory Opinion from 19 July 2024, the International Court of Justice declared the Israeli occupation of Palestine as unlawful under international law and found a series of violations against IHRL in a form of, for instance, forcible evictions and discriminatory practices against the Palestinian population. The UN Special Rapporteur on the situation of human rights in the Palestinian territories occupied since 1967, alongside other experts and civil society, has repeatedly emphasized that the Israeli offensive in Gaza meets the criteria for being classified as genocide.

In 2025, owing to mounting pressure from certain Member States and EU citizens, the European Commission has proposed a series of measures in response to Israel’s violations of international law in the occupied territories of Palestine. The High Representative of the EU for Foreign Affairs and Security Policy announced a review of the EU-Israel Association Agreement in order to asses Israel’s compliance with human rights and democratic principles (link). Consequently, the Commission proposed a partial suspension of Israel’s association to Horizon Europe (link). Moreover, it put forward a proposition to suspend trade concessions with Israel and to impose sanction on certain individuals (link). However, at the time of finalizing this article in January 2026, no significant sanctions had been imposed by the European Union itself.

In the light of the EU’s insufficient response to the breaches of international law, JURDI, the Association of Jurists for the Respect of International Law, founded in 2024 in order to promote respect for and the application of international law in the context of the Israeli-Palestinian conflict (link), called the EU institutions earlier this year to take appropriate action to ensure respect for international law. In the absence of appropriate response to the request, JURDI lodged a complaint under Article 265 TFEU for the failure to act.

Procedural pitfall

The GC declared the action manifestly inadmissible without further examining its merits, because it found that the applicant lacked standing under Article 265 TFEU. In that respect, GC reminded that Articles 263 and 265 TFEU constitute the expression of a single legal remedy, which means that conditions to attribute standing to an individual under 265 TFEU are virtually the same as those prescribed in 263 TFEU (para 10). Meaning that individuals can bring an action for annulment against an inaction of an institution not addressed to them, provided that the act concerns them directly and individually, or against regulatory acts that concern them directly and that do not include implementing measures (mutatis mutandis para. 10). Additionally, according to the CJEU’s case law, associations can generally be attributed legal standing if: (1) a legal provision expressly recognizes a series of procedural powers for such associations; (2) association represents the interests of individuals which would be entitled to bring an action, (3) the association is individualized because of the impact of the measure on its own interests as an association (see here and here).

Despite the claimant’s arguments, GC didn’t find it directly and individually affected by the inaction of the Council and the Commission. In that respect, it observed that none of the three measures that the Council and the Commission allegedly failed to adopt would directly affect JURDI’s legal situation by modifying its rights or obligations (para. 13). Especially, GC did not agree with the applicant’s argument that it was concerned directly since its statutory mission was based on the prevention of international crimes and genocide (para. 14, following Union syndicale-Service public européen ea v Council case-law). GC also found that the measures mentioned in the claim would not produce any direct effect on the legal situation of the JURDI’s members (para. 17). Additionally, GC established that JURDI is not individually concerned, since there is no clear evidence that it can be distinguished from any other association charged with similar mission (“contributing to the promotion, defence, respect and implementation of international law in France, Europe and the world in the context of the Israeli-Palestinian conflict”) (para. 20).

The GC didn’t also accept JURDI’s argumentation based on ClientEarth case in which a non-governmental organisation successfully brought an action for annulment in a public interest against a decision of the European Investment Bank rejecting the applicant’s request for an internal review of the resolution of the EIB’s Board of Directors approving the financing of a biomass power generation plant in Galicia. According to the GC, situations in these cases were not comparable, since in ClientEarth the Court simply recognized that an environmental NGO was entitled to request such a review in accordance with Regulation (EC) No 1367/2006 of the European Parliament and of the Council of 6 September 2006 on the application to the institutions and bodies of the European Union of the provisions of the Aarhus Convention on Access to Information, Public Participation in Decision-making and Access to Justice in Environmental Matters (OJ 2006, L 264, p. 13). Consequently, in ClientEarth the Court therefore did not rule on the admissibility under the general conditions laid in Article 263 TFEU but only applied an existing act of secondary legislation.

Criticism against the CJEU’s rigid stance

The JURDI case is yet another example of failed public interest litigation, which couldn’t stand the rigidity of the accessibility test under Articles 263 and 265 TFEU. The issue of the rigid standing criteria limiting, in many instances, access of individuals to the CJEU has been broadly discussed in the literature (see for instance here and here). It has also been a matter of consideration in prominent advocate general’s opinions both before the adoption of the Lisbon Treaty (see opinion in Unión de Pequeños Agricultore) and more recently in AG Emiliou’s opinion given in Nicoventures.

AG Emiliou, in his opinion, pointed out two major strands of criticism against the interpretative choices of the CJEU that led to a narrow understanding of the terms “direct” and especially “individual” concern. The first strand of criticism highlights that the case law of the CJEU on admissibility is complicated and incoherent. For example, in the case of locus standi of international actors, the CJEU tends to be more flexible in the interpretation of the premise of “legal person” mentioned in Article 263 TFEU (see discussion here). The second group of arguments concentrates on the fact that the current interpretation of the standing criteria is rendering public litigation in the common interest practically impossible in many instances (like in the case of legislative acts). Indeed, the recent Caravhalo judgement depicts how reluctant the CJEU is to relax its criteria of accessibility even in cases concerning the common threat that climate change poses.

As AG Emiliou rightly pointed out, it stands in contradiction with the ambitious axiological foundations of the Treaties (see the Opinion critically discussed here). Indeed, the strategic litigation of NGOs commonly concerns the protection of EU values like human rights and the rule of law. That was the case of the JURDI’s application. The GC was particularly silent on the issue that the goals the association pursues reflect the very values prescribed in Article 2 TEU as well as the principles and objectives of the EU’s external action recognized in Articles 3 (5) and 21 TEU. In that respect the GC merely followed the case law established since Inuit Tapiriit Kanatami – interpretation of the Article 263 TFEU cannot set aside conditions lied in Treaties of direct and individual concern (para. 27). However, one could wonder whether the CJEU really cannot make an interpretative choice to understand an association as directly and individually concerned by EU’s inaction that leads to the undermining of said associations mission (preventing genocide).

The argument for the admissibility of the public interest litigation

One can argue that the system of legal remedies in the EU is generally aimed at protecting individuals whose rights and freedoms are guaranteed by EU law, and have been violated, as stated in Article 47 of the EU Charter of Fundamental Rights. Thus, actions for annulment and for failure to act must be understood as measures designed to protect individual interests of the applicants. This would imply that since neither JURDI nor its members are affected by the institutions’ failure to act, the organisation cannot successfully initiate court proceedings. However, such a narrow reading of the provisions of the Treaties does not seem supported by the fundamental values and principles of the EU that should guide the Court in its interpretative activity. The Court has previously demonstrated its capacity to modify the interpretation of the Treaties when overriding considerations necessitate such action, as evidenced in its ruling in Les Verts. Why, then, wouldn’t the CJEU reconsider its stance on the criteria for admissibility in the context of Articles 263 and 265 TFEU?

First, as stated before, it can be argued that a civil society organisation should have a right to an effective judicial protection whenever an action or a failure to act on part of EU institutions threatens values which the organisation is promoting and protecting. This, in fact, constitutes an individual concern since ensuring respect for these values is the essence of the existence of such an organisation.

Second, one may question the Court’s stance on admissibility in light of the fact that the EU is a community based on the rule of law. However, a rigid interpretation of Article 263 and 265 TFEU has the effect of reducing the ability of this community to ensure that the law is upheld. Indeed, it reinforces a system of self-policing, which may be regarded as questionable not only from the perspective of rule of law but also the democratic foundations of the Union. It is, therefore, imperative that the CJEU, as the custodian of the law, assumes the responsibility entrusted to it by the Treaties (Article 19 TEU) to ensure compliance with fundamental values and, in order to fulfil this obligation, grant civil society organisations the possibility of litigation in the public interest. The question of admissibility does not prejudge the merits of the complaint. It merely establishes a forum through which compliance with fundamental values can be ensured, especially where neither EU institutions themselves nor Member States exhibit concern for the matter.

The additional dimension of this case that underscores the CJEU’s obscure reluctance is the fact that the EU-Israel agreement contains a human rights clause. Provisions designed by the EU precisely to avoid the situation of engagement in trade relations with a country that is violating international law. Suffice to say, the inspiration for its adoption was the suspension of the agreement with Yugoslavia in 1991 based on the customary international law (Racke). Human rights clauses have been mainstreamed for years on the EU’s own initiative.

Conclusions

Is public interest litigation before the CJEU a Sisyphean task? Despite the fact that the Court’s current case-law does not instil a sense of optimism, we remain hopeful that the answer to this question is a negative one. We believe it to be a laborious task, but not a futile one. Why? The current framework of the Treaties does not preclude the admissibility of the actions for annulment and for failure to act brought by civil society organisations in the public interest – only, in many instances, the interpretation given to it by the CJEU does. However, throughout the course of its history, the CJEU has repeatedly demonstrated its capacity to interpret acts of EU law from a liberal and pro-value standpoint.

In fact, it is worthwhile to critically revisit in this context the GC’s unfortunate and blatant observation that “associations have no right, under the Union legal order, to have the objectives of their activities unaffected by acts of the Union” (para. 16). Such an unnuanced reflection does not change anything as regards procedural law. It seems, however, out of kilter and in contrast with EU’s axiological foundations in the context of JURDI’s objectives to prevent international crimes and genocide.

The post Time to Rethink Locus Standi appeared first on Verfassungsblog.