By Mona Grimm and Lena Riemer
Last month, Austria signed its new migration agreement with Uzbekistan, quietly and without much fanfare – so little fanfare even, that its text remained unknown to the public for several weeks. When the agreement finally became available, it stood out that it was presented as a balanced framework for mobility, skilled migration, and cooperation on return. Yet, beneath the language of partnership lies a broader transformation in European migration governance, in which the return responsibility for deportable citizens and non-citizens is increasingly externalised through bilateral frameworks. In the case at hand, the agreement inter alia enables the rapid transfer of deportable Uzbek citizens, stateless and non-citizens, including rejected asylum seekers from Austria to Uzbekistan, which has agreed to take them back (own citizens) or organise their onward ‘transfer’ to their respective countries of origin (Durchbeförderung).
In doing so, the Austria–Uzbekistan agreement demonstrates how this emerging pattern of externalised return is imagined by states to work in practice. While it opens selective pathways for highly skilled workers, researchers, and students, its operational core is devoted to readmission and return. On the macro level, the agreement reflects a wider European shift away from relational understandings of return between hosting state and state of origin toward a system in which transit history itself becomes sufficient to allocate responsibility for receiving deported individuals such as rejected asylum seekers.
This post examines the Austria–Uzbekistan agreement in detail and argues that it should be understood as part of a structural reconfiguration of European asylum and return governance. The agreement mirrors broader developments under the EU Pact on Migration and Asylum, and the proposed Return Hubs, including responsibility shifting for protection and return through procedural mechanisms, external partnerships, and weakened requirements of meaningful connection to third countries.
The Austria–Uzbekistan Agreement: Selective Mobility and Expanded Return Governance
The wording of the agreement itself already indicates its political orientation. Throughout the text, Uzbekistan appears as the “requested contracting party” (“ersuchte Vertragspartei”), revealing Austria as the initiating actor for the return of non-citizens. This aligns directly with the Austrian government programme 2025–2029, which explicitly envisages “strategic partnerships” with third countries as instruments for migration control and the prevention of irregular migration. Statements by Austrian officials, including Foreign Minister Beate Meinl-Reisinger, frame the agreement primarily as a tool to address irregular migration.
At first glance, however, the agreement presents itself as balanced cooperation. The first nine articles establish selective mobility pathways for Uzbek professionals, students, and researchers. Like many contemporary European migration partnerships, the agreement combines labour migration objectives with return cooperation. It thereby links access to mobility opportunities with Uzbekistan’s’ willingness to facilitate deportation of their own nationals, and in some cases third-country nationals, who do not, or not longer, meet the conditions for residence in Austria. The structure of the agreement reveals a clear asymmetry. While the provisions on legal migration remain relatively general, the substantive core of the agreement is devoted to the deportation and the readmission of Uzbek citizens, as well as third-country nationals and stateless persons with minimal links to Uzbekistan, to latter’s territory. Articles 11–24 occupy significantly more space and provide detailed operational mechanisms governing return procedures, documentation, transit, identification, deadlines, and cooperation obligations. The agreement’s core therefore lies in creating efficient pathways for deportation.
This asymmetry reflects a wider trajectory we can currently witness in the field of European migration governance. Across the EU, mobility is increasingly selective and economically conditional, privileging highly skilled migrants, students, and researchers, while return infrastructures are simultaneously expanded and accelerated. A prominent example in this regard is the currently proposed EU ‘return hubs’ envisioning detention and processing facilities in non-EU countries. Designed to hold individuals with rejected asylum claims or those residing unlawfully in the EU while they await deportation, this model of externalization has been criticized for raising questions of fundamental guarantees such as non-refoulement. The Austria–Uzbekistan agreement reproduces this logic of speedily removing non-citizens and stateless individuals as efficiently as possible to third countries who are willing to take them in.
Particularly significant is the agreement’s inclusion of third-country nationals and stateless people within its deportation framework. Article 13 of the agreement extends readmission obligations beyond Uzbek nationals, creating the possibility that individuals may be transferred to Uzbekistan despite lacking meaningful social or territorial ties to the country. Under Article 13 paragraph 3 (a), merely transitioning through Uzbek territory establishes a sufficient nexus for a potential deportation. Additionally, Appendixes 3 and 4 introduce low-certainty evidence such as witness statements or statements from Uzbek officials to attest an individual’s stay or transit in Uzbekistan. These evidentiary standards could potentially create vulnerabilities regarding misidentification and non-refoulement obligations. This agreement, therefore, marks an important development in Austria’s bilateral migration policy by expanding its deportation framework to third-country nationals and stateless persons who lack meaningful ties to Uzbekistan. While Austria previously concluded comparable agreements, including a 2023 migration partnership with India, the Uzbekistan agreement introduces a qualitatively broader framework through its inclusion of the return of third-country nationals.
When we move from the domestic to the EU level, we see a pattern emerge. This development strongly mirrors and builds on ongoing transformations at the EU level. Under the New Pact on Migration and Asylum, the traditional “link requirement” for safe third country concepts is significantly weakened. Historically, inadmissibility decisions depended upon meaningful connections between an applicant and the third country in question, such as prior residence, family ties, or cultural connections. Increasingly, however, transit alone may become sufficient to justify transfer or inadmissibility, especially where cooperation agreements exist.
The Austria–Uzbekistan agreement operationalises this logic. Responsibility is no longer allocated according to substantive connections, transiting a country may now be sufficient for the individuals’ return thereto. The political context surrounding the agreement reinforces this interpretation. Austrian politician and Member of the European Parliament Reinhold Lopatka has explicitly referred to Uzbekistan’s geographical proximity to Afghanistan, to which returns are legally and politically at least difficult, as a factor underlying such migration partnerships. Austria thereby follows a path already pursued by Germany, which concluded a similar agreement with Uzbekistan in 2023. Media reporting surrounding both the German and the Austrian agreement has fueled speculation that these return arrangements may indirectly target Afghan nationals.
Efficiency over Safeguards? The Risks of Violating Human Rights in the Austria–Uzbekistan Agreement
Even beyond questions of political targeting, the agreement raises serious concerns regarding procedural safeguards and the protection of individual rights. For example, Article 15 governs the determination of nationality and identity. Notably, paragraph 4 permits virtual identification procedures, allowing nationality assessments to occur remotely. While maybe procedurally efficient, such methods create substantial risks of misidentification, especially in complex cases involving incomplete documentation, language barriers, or disputed nationality claims.
These risks increase by the agreement’s accelerated timelines. The requested state must respond to identification requests within 30 calendar days (Article 15 (5)), while travel documents are to be issued within seven days (Article 15 (6)). Such compressed procedures leave limited room for factual correction, legal intervention, or effective challenge. The consequences may be particularly severe in the context of non-refoulement obligations. Erroneous nationality determinations or incorrect assumptions regarding connection to Uzbekistan could expose individuals to onward removal, persecution, or other forms of serious harm in Uzbekistan or in case of onward transfer. Although Article 19 provides for the correction of wrongful deportations, the safeguard is significantly weakened by a limitation period. Austria is obliged to readmit wrongly transferred individuals only within six months following transfer. After that point, corrective responsibility effectively expires. Such temporal limitations are deeply problematic where human rights violations may emerge only after prolonged detention, onward removal, or deteriorating conditions in the receiving state. The structure of the agreement, therefore, risks reducing protection guarantees to narrow procedural formalities rather than ensuring substantive safeguards against refoulement.
In addition, and as so often the issue with ‘voluntary return’ options, here too, the agreement’s reliance on this concept for the group of Uzbek citizens raises the question on the thin line between formal voluntariness and structural ‘coercion’. Article 12 explicitly promotes voluntary return mechanisms for this group. Yet, because consent is not required for deportation under paragraph 2, the voluntariness of such returns exists within a framework shaped by the credible threat of forced removal. Interestingly, a similar clause regarding voluntariness cannot be found in relation to deporting third-country nationals and stateless persons. What is true for all groups listed in the agreement as potentially affected by this return agreement are serious concerns on the human rights situations in Uzbekistan itself – not to even mention the situation in case non-citizens were to be transferred onward to Afghanistan. Recent reports by organisations such as Amnesty International and Human Rights Watch have continuously documented restrictions on freedom of expression, freedom of religion, and LGBTQIA+ rights in Uzbekistan. These concerns become particularly significant in the context of accelerated deportation procedures and limited corrective safeguards and monitoring.
Conclusion – The Transformation of European Migration Governance
As this analysis has shown, the recent Austria–Uzbekistan agreement is more than a bilateral migration arrangement. It illustrates a broader transformation in European migration governance, in which the externalisation of return increasingly occurs through legal architecture, procedural acceleration, and strategic partnerships with third countries. This recent bilateral agreement reflects wider developments currently reshaping EU asylum and return policy. The agreement contributes to the erosion of the traditional ‘link requirement’ in asylum and return law. Responsibility for the non-citizen is no longer grounded in meaningful ties, making transit itself sufficient for removal. Besides that, the agreement demonstrates how contemporary externalisation increasingly operates through legal and procedural design rather than physical outsourcing alone. Tight deadlines, remote identification procedures, limited corrective safeguards, and weak suspensive remedies prioritise efficient transfer.
In this sense, in these weeks leading up to both the entry into practice of the New Pact on Migration and Asylum and the last push for EU return hubs, the Austria–Uzbekistan agreement should be understood as part of a wider European shift toward governing migration through externalised responsibility-sharing frameworks. The emerging system increasingly focuses on how ‘undesired’ individuals can be returned elsewhere through pre-arranged cooperation mechanisms. As the EU Pact on Migration and Asylum and proposed ‘return hubs’ continue to expand this logic, bilateral agreements such as the Austria–Uzbekistan one offer an important preview of what may may become a common standard of return externalisation based on agreements with third countries outside of the European Union with little or no link to the deportable individual.
Lena Riemer, LL.M. (Yale) is an Assistant Professor of law at Central European University, Vienna. Her work focuses on human rights, climate change and refugee and migration law.
Mona Grimm is a student at Central European University, where she works with Dr. Lena Riemer as a research assistant. She has contributed to projects on expulsion policies, externalization, and an amicus curiae brief on climate mobility.
The views expressed in this article belong to the author/s and do not necessarily reflect those of the Refugee Law Initiative. We welcome comments and contributions to this blog – please comment below and see here for contribution guidelines.