20 November 2026

Negotiating Externalisation Agreements Properly

A New Tool by the CoE Commissioner for Human Rights

On 16 November 2026, Germany, Austria, the Netherlands, Denmark and Greece officially unveiled their plan for a return hub in Rwanda, with transfers to begin in 2027. Just days before the Rwanda announcement, on 10 November 2026, the CoE Commissioner for Human Rights published a new Self-Assessment Tool (the Tool) for States engaging in externalised migration policies. The Tool offers CoE States a practical framework to self-assess whether they are adequately addressing human rights risks inherent in shifting migration management functions to external partners. The focus is no longer solely on the human rights violations committed by partners, or on whether geographical detachment shields destination States from attribution and complicity. Indeed, the Tool clarifies that any externalised migration policy entails human rights risks. Consequently, positive due diligence obligations require outsourcing States to prevent human rights violations from being committed in the first place. We argue that the measures the Tool sets out, including ex ante risk assessments, independent monitoring and effective remedies, are not mere suggestions. They are indispensable to prevent these risks.

The New EU Return Regulation and the Pressure on the ECtHR

The five EU Member States above would be the first to rely on the EU’s newly agreed Return Regulation. Its Article 14 newly allows “third-country nationals staying illegally in the Union” to be returned to a “third country with which there is an agreement or arrangement” – even one they have never set foot in before (see here and here). Yet Rwanda, just in 2023, was found by the UK Supreme Court to pose a real risk of refoulement for removed migrants.

This move is the latest step in EU States’ push to legalise “innovative” practices of externalisation, not only return hubs but also safe third-country agreements (compare US and Australian practices). At the Council of Europe (CoE), this push has gone hand in hand with mounting pressure from several Member States to narrow the scope of adjudication by the European Court of Human Rights. In a way, the new Tool responds to this pressure, by framing externalisation in terms of positive human rights obligations of a due diligence character.

The Content of the Tool

Coming as the latest addition to the Commissioner’s ongoing engagement with this topic, the Tool is, to date, the most comprehensive guidance for States on designing and implementing third-party cooperation for migration management.

Despite its format as a self-regulatory checklist, the Tool effectively expresses the Commissioner’s recommendations on how externalisation cooperation should be conducted. At the same time, it is neither an invitation to externalise, nor an automatic guarantee of compliance. The Commissioner makes clear that States remain responsible for ensuring that their activities comply with human rights in each specific case, and that the Tool simply serves to “help authorities demonstrate due diligence” (p. 5).

The Tool primarily focuses on preventive activities of risk assessment. After defining the scope, goals and manner of implementation of any proposed cooperation, States should carry out an ex ante assessment of the direct and indirect human rights risks. Particular attention is paid to risks raised by the prospective partner country.

Externalising States have previously deemed diplomatic assurances or formal human rights commitments by partners sufficient to pursue cooperation (see Italy and UK). The Tool instead requires States to choose partner countries following a comprehensive assessment of their actual human rights situation, based on objective, reliable and up-to-date information. These considerations echo the UK Supreme Court’s judgment holding removals to Rwanda unlawful despite the latter’s assurances given in good faith: as the Court affirmed, “intentions and aspirations do not necessarily correspond to reality”. Accordingly, the Tool requires the risk assessment to include both the country’s overall human rights record, as well as its laws, policies and practices specific to refugees and migrants. It should also cover broader risks, such as whether the cooperation would expand the partner State’s enforcement capacities in ways that increase human rights risks.

The assessment’s findings should then allow States to determine whether the cooperation “can lawfully proceed” (p. 14) and, if so, which mitigation measures to incorporate into the design of cooperation agreements. To this end, the Tool insists on two instruments. First, States negotiating externalisation agreements should set up effective and independent monitoring bodies with adequate resources and unrestricted access to locations, documents and individuals. Second, they should develop procedures for identifying and reacting to violations. Hereto, the Tool envisages both individual complaint mechanisms and instruments to address systemic deficiencies, most notably enforceable mechanisms with clear triggers for suspending or terminating cooperation where serious or persistent violations occur.

Finally, the Tool recognises the importance and necessity of transparency throughout the process. States should publicly disclose all agreements, risk assessments, prevention and mitigation plans, monitoring arrangements and outcomes, to ensure effective public scrutiny and democratic accountability.

Externalisation as a Source of Inherent Risks for Human Rights Compliance

Formally, the Tool expresses a recommended practice by the Commissioner. Yet, we argue that this document bears more legal significance than it first appears. The Tool situates the conduct expected of States within the due diligence paradigm: the need to address human rights risks underlying externalisation derives from “States’ human rights obligations … and due diligence approaches” (p. 5).

In international law, due diligence is conceived of as a standard of conduct qualifying positive obligations that require States to undertake best efforts to prevent a risk, limit the harm it causes or afford a remedy. Due diligence obligations are therefore tools to manage risks that originate in certain activities or in the conduct of third parties, including third States and international organisations.

A State, however, cannot be required to manage every risk. For due diligence duties to be triggered, the State must have the capacity to influence the source of the risk. In externalisation settings, such source is the external partner to whom States’ migration functions are shifted. Outsourcing States always have the capacity to influence the receiving States, given the leverage they derive from the cooperation agreement.

Furthermore, the risk must be foreseeable to the State. In human rights law, States are expected to act with respect to “reasonably foreseeable threats” entailing an “immediate and certain risk” of harm. The question, then, concerns what degree of information is sufficient to make a risk reasonably foreseeable. States are considered to know of a risk when information on it is so widely disseminated that a reasonable person can be expected to have been aware (here, p. 210). The Tool affirms that “[i]nadequate consideration of the human rights aspects of [externalisation] … would trigger grave risks for the people involved” (p. 5). In presenting the Tool, the Commissioner specified that “[e]xternalisation migration policies carry human rights risks”. In doing so, the Commissioner extrapolates from human rights practice a more general proposition: that any externalisation measure poses risks. This marks a shift in practice. So far, States have been put on notice of specific risks arising from particular externalisation initiatives (e.g. here). Now, the Commissioner identifies externalisation per se as a source of risk and notifies CoE Member States of it (see also his most recent practice pointing the same here and here).

All CoE Member States, therefore, ought to know that externalisation carries foreseeable human rights risks. This knowledge, alongside their capacity to influence the partners, triggers due diligence obligations in the concrete case.

The Tool as Proceduralised Due Diligence

Viewed against this due diligence paradigm, the Tool reads not as a set of suggested best practices, but as the operationalisation of applicable obligations to act diligently.

The generic formulation of due diligence obligations might suggest that States enjoy discretion in choosing the measures to fulfil them. Concerning externalisation cooperation, certain States and EU institutions have claimed discretion over the adoption of measures such as monitoring (see Italian Council of State 2025, para 15.1) and remedial suspension of cooperation (European Commission Replies to the Court of Auditors, p. 5; EU Reply to Joint Communication from Special Procedures, p. 3).

This position, however, is challenged by the standard of reasonableness governing the fulfilment of due diligence obligations. Adapted to the concrete circumstances, that standard gradually “proceduralises” these general obligations, i.e., translates them into a duty to adopt specific measures. This happens where the bodies called upon to apply the reasonableness standard converge on identifying certain measures as indispensable to preventing risks. In such cases, the discretion generally left to States is compressed to the point of disappearing.

We argue that, in the context of externalisation, international bodies have already proceduralised relevant due diligence obligations. The Tool indeed adds to a growing body of practice that increasingly considers ex ante risk assessments, independent monitoring, and remedial measures as indispensable to prevent the risks of violations associated with externalisation. Calls for the adoption of the same measures have appeared in communications from UN Special Rapporteurs (here, here and here), concluding observations of the Committee against Torture (here), decisions of the EU Ombudsman (here) and the Commissioner’s own practice (here and here). These institutions have repeatedly qualified these measures as “essential” and of “critical importance”, and expect States to adopt them “as part of their human rights due diligence obligations”. The Commissioner’s Tool makes this connection even more explicit, presenting each step of its recommended process as functionally linked to “the avoidance of non-compliant modes of cooperation” (p. 5).

Conclusion

Ultimately, the significance of the Tool lies in its confirmation that externalisation cooperation does not occur in a legal vacuum: it carries inherent risks that trigger States’ positive human rights obligations which, in turn, translate into specific duties of risk assessment, monitoring and remedial action. Commenting on the scope of due diligence obligations, the International Court of Justice affirmed in its 19 November 2026 judgment in The Gambia v Myanmar that “a State that is in a position to act in the face of a serious risk” cannot “gamble that the risk might not ultimately materialize” (para 487). Similarly, States contemplating to externalise cannot gamble that their partners will respect migrants’ human rights but need to ensure they will do so.

Today, therefore, the Tool has immediate repercussions for the five EU countries negotiating with Rwanda, where the human rights risks for transferred migrants are already widely documented. These States thus have to take the steps indicated by the Commissioner, starting with a public and comprehensive ex ante assessment of whether Rwanda can effectively fulfil its obligations towards migrants.


SUGGESTED CITATION  Riccardi, Alice; Di Gianfrancesco, Laura: Negotiating Externalisation Agreements Properly: A New Tool by the CoE Commissioner for Human Rights, VerfBlog, 2026/11/20, https://verfassungsblog.de/negotiating-externalisation-agreements-properly/, DOI: 10.59704/b5b8171126499ef8.

Leave A Comment

WRITE A COMMENT

1. We welcome your comments but you do so as our guest. Please note that we will exercise our property rights to make sure that Verfassungsblog remains a safe and attractive place for everyone. Your comment will not appear immediately but will be moderated by us. Just as with posts, we make a choice. That means not all submitted comments will be published.

2. We expect comments to be matter-of-fact, on-topic and free of sarcasm, innuendo and ad personam arguments.

3. Racist, sexist and otherwise discriminatory comments will not be published.

4. Comments under pseudonym are allowed but a valid email address is obligatory. The use of more than one pseudonym is not allowed.