Monday, 21 April 2025

Jumping the Gun? The proposed early application of some of the EU’s new asylum pact – and a common list of supposedly ‘safe countries of origin’

 



Steve Peers, Professor of Law, Royal Holloway University of London

Photo credit: Andre Engels, via Wikimedia Commons

The EU’s asylum pact was adopted less a year ago, and mostly won’t apply for over another year – and yet the EU Commission has already proposed to amend it, in order to bring forward some of the rules in the procedural part of the pact, and to adopt a common list of ‘safe countries of origin’ to apply when the rest of the pact enters into force. The aim is to speed up consideration of asylum claims, and in particular to help to ‘save’ the Italy/Albania deal on asylum processing. The following blog post looks in turn at the background to the new proposal, and then the different elements of it, followed by an assessment.  

Background

Previous and current rules

Initially, the concept of ‘safe countries of origin’ goes back, at EU-wide level, to ‘soft law’ adopted in the early 1990s (one of the ‘London Resolutions’ of 1992). Subsequently, the principle took on binding legal form at EU level in the first-phase 2005 asylum procedures Directive, which provided for an option for Member States to accelerate considering asylum applications (albeit in accordance with the usual procedural rules), inter alia where the applicant is from a ‘safe country of origin’, as further defined. (Note that these rules refer to non-EU countries of origin; there is a separate, stricter set of rules setting out the near-impossibility of EU citizens making asylum applications in other Member States, because each EU Member State is considered to be a ‘safe country of origin’ too, according to a protocol attached to the EU Treaties).

Currently, a revised version of the principle is set out in the second-phase asylum procedures Directive, adopted in 2013 (the ‘2013 Directive’). Unlike the 2005 Directive, there is no longer a reference to potentially treating only part of a country as ‘safe’, and the previous option for Member States to retain pre-existing lower standards on this issue (along with pre-existing rules on designating part of a country as ‘safe’, or as ‘safe’ for groups of people) was dropped.

The CJEU has ruled on these provisions twice. First, the Court confirmed that Member States had to provide for a ‘safe countries of origin’ rule in national law if they wanted to use apply this principle. Secondly, in October 2024 the Court interpreted the substance of the rule, in particular confirming that it was no longer possible to designate part of country of origin as ‘safe’, given that the EU legislator had dropped that possibility from the text of the 2013 Directive, as compared to the 2005 Directive (see further discussion of that judgment here).

Given that the Italy/Albania treaty on housing asylum applicants in Albania only applied (at least initially) to asylum-seekers from supposed ‘safe countries of origin’, this created a number of potential barriers to the application of that treaty, with multiple Italian courts sending a questions to the CJEU about the rule. The CJEU has fast-tracked two of these cases – Alace and Canpelli – which raise questions in particular about whether Member States can designate a country of origin as ‘safe’ with exceptions for certain groups, and also whether they can designate such countries by means of legislation and must publish the sources of their assessment when they do so. (The case is pending: see earlier blog posts on the background, the hearing, and the Advocate-General’s opinion)

Future rules

The 2024 asylum procedures Regulation (the ‘2024 Regulation’) has amended the ‘safe country of origin’ rules again, although as things stand the 2024 Regulation is only applicable to applications made after June 2026. This upcoming version retains many of the current features of the ‘safe country of origin’ concept (which are set out in more detail below): the definition of human rights standards which must apply before a country can be designated as ‘safe’; the procedure for designation (laying out the sources of information which must be taken into account); and the safeguards (the asylum-seeker must be a national of or a stateless person habitually resident in the country concerned, and must have the possibility to rebut the presumption of safety in their particular circumstances).

But there are several changes in the 2024 Regulation. In particular, it will now again expressly be possible to create an exception to the designation of ‘safety’ for ‘specific parts’ of the non-EU country’s territory and (not only in the context of pre-existing law) for ‘clearly identifiable categories of persons’.

More broadly, the ‘safe country of origin’ rule will remain on the list of possible accelerated procedures, but there is more harmonisation of the rules on time limits and appeals in these cases. There is also a potentially overlapping new ground of accelerated procedures where the country of origin has an international protection recognition rate below 20% at first instance (based on the latest annual Eurostat data), although this is subject to some safeguards, discussed further below.

Another important new development in the 2024 Regulation is the possibility to adopt a common EU list of ‘safe countries of origin’ (there were two earlier failed attempts to do this; see my previous blog post). According to Article 62(1) of the Regulation in its current form, the EU common list must be subject to the same rules as the national list (‘in accordance with the conditions laid down in Article 61’). The Commission has to review the EU list with the assistance of the EU Asylum Agency, on the basis of the sources of information applicable to Member States drawing up their lists (Article 62(2)). Also, the EU Asylum Agency must provide information to the Commission when it draws up proposals for the common EU list (Article 62(3); the list must be adopted by the ordinary legislative procedure, ie a qualified majority of Member States, in agreement with the European Parliament). If there are ‘significant changes’ in a country on the common EU list, the Commission must conduct a ‘substantiated assessment’ of the situation in light of the ‘safe country of origin’ criteria, and can suspend a country from the list on a fast-track basis.

As for Member States, they can still designate additional countries as ‘safe countries of origin’, even if those countries are not on the common EU list. But if a country is suspended from the common EU list, Member States need the Commission’s approval to put that country back on a national list for the following two years.

The new proposal

The new proposal has two main elements, each of which can be broken down into two sub-elements. First of all, it would bring forward some of the rules in the 2024 Regulation. This would apply to aspects of the ‘safe country of origin’ and ‘safe third country’ rules on the one hand (which would apply when the newly proposed Regulation, once adopted, enters into force), and to the ‘low recognition rate’ ground of accelerated proceedings on the other (which Member States could apply before the asylum pact otherwise applies).

Secondly, it would establish a common EU list of ‘safe countries of origin’ that would apply as from the main 2026 date to apply the 2024 Regulation as a whole. This would include both candidate countries for accession to the EU (which would be subject to a new set of special rules) and a further list of seven countries to be regarded as ‘safe countries of origin’.

The proposal would apply to all Member States except Denmark and possibly Ireland, which could opt in or out (so far, Ireland has adopted into all of the asylum pact measures that it could).  It would not apply to non-EU countries associated with Schengen.

Earlier application of the asylum pact

‘Safe country’ rules

The proposal would allow the earlier application of key changes to the ‘safe country of origin’ rules set out in the 2024 Regulation, as regards creating exceptions to that concept for part of a country, and for groups of people. As noted above, the CJEU has ruled that the former exception cannot apply under the 2013 Directive, while it will soon rule on whether the latter exception can currently be invoked under that Directive. So if the proposal is adopted, the change as regards exceptions for part of a country will definitely overturn the existing case law, while the change as regards exceptions for a group of people will possibly change the existing law, depending on what the Court rules (it’s likely, but not certain, that the judgment will come before the proposal becomes law).

Of course, these changes will apply anyway once the 2024 Regulation applies in June 2026. But some Member States are anxious to be able to apply these exceptions earlier than that, in particular Italy: both the exceptions are very relevant in practice to whether the Italy/Albania asylum deal is workable earlier than next June.

The proposal would also allow the earlier application of the same changes to the ‘safe third country’ rules set out in the 2024 Regulation (ie the rules on whether asylum seekers can be sent to another country, other than an EU Member State or their country of origin, which should decide upon their asylum application). Presumably the Commission assumes that the CJEU, if asked, would also find that there is no exception for parts of a country or groups of people as regards designation of ‘safe third countries’, by analogy with its existing or possible future judgments on ‘safe countries of origin’ under the 2013 Directive.  

Note that only some of the new ‘safe third country’ and ‘safe country of origin’ rules in the 2024 Regulation (ie the possible exceptions for parts of countries or groups of people) would apply early. For instance, the prospect of common EU lists for either concept would not apply early; the proposed common ‘safe country of origin’ list, discussed below, would only apply from June 2026, when the 2024 Regulation generally starts to apply. Furthermore, the Commission will likely soon propose further changes to the ‘safe third country’ rules, in a separate proposal: the 2024 Regulation requires a review of those rules by this June.

Low recognition rate rules

In addition to early application of revised versions of current rules, the proposal would also bring forward the application of a brand new rule set out in the 2024 Regulation: the ‘low recognition rate’ rule, on accelerated procedures where the recognition rate (ie the success rate of asylum applications) of a country’s citizens is below 20% at first instance, ie before appeals (even though a proportion of appeals is successful). This also includes most of the safeguards attached to this new rule: it cannot apply if the Member States’ administration ‘assesses that a significant change has occurred in the third country concerned since the publication of the relevant Eurostat data or that the applicant belongs to a category of persons for whom the proportion of 20 % or lower cannot be considered to be representative for their protection needs, taking into account, inter alia, the significant differences between first instance and final decisions’.

The proposal also provides for early application of the same rule (subject to the same safeguards) as regards unaccompanied minors, although the Commission makes no mention of this point, and so provides no justification for it, in its explanatory memorandum.

However, arguably the proposal does not bring forward the rule (as regards both asylum seekers generally and unaccompanied minors in particular) that the assessment of significant changes must take account of any guidance note on the point issued by the EU Asylum Agency.

Also, the proposal does not bring forward other aspects of the 2024 Regulation related to the ‘low recognition rate’ rule. The Commission expressly points out that the rule will remain optional for Member States, until the 2024 Regulation makes it mandatory from June 2026. Furthermore, while the proposal states that the ‘low recognition rate’ rule can be used in special border procedures (in the 2013 Directive version of border procedures, not the 2024 Regulation version of them, until June 2026), it does not include the important exceptions from border procedures set out in the 2024 Regulation.

In particular, that Regulation excludes the border procedure from applying to unaccompanied minors on ‘low recognition rate’ grounds, and also excludes the border procedure from applying to asylum seekers generally where: the rules on accelerated or inadmissible cases do not apply; support cannot be provided to asylum seekers with ‘special reception needs’ or ‘in need of special procedural guarantees’; there are medical grounds; or detention guarantees cannot be complied with. But none of these exceptions are made applicable (prior to June 2026) by the new proposal. This point is particularly relevant to detaining asylum seekers – which is easier to justify legally when the border procedure applies. So the attempt to widen the use of the borders procedure could widen the use of detention.

Common EU list of ‘safe countries of origin’

EU accession candidates

The proposed Regulation would delete the current Article 62(1) of the 2024 Regulation (which requires any common EU list of ‘safe countries of origin’ to comply with the ‘conditions’ relating to that concept set out in Article 61), replacing it with a statement that candidate countries to join the EU (the Commission does not name them, but they are Serbia, Montenegro, Ukraine, Moldova, North Macedonia, Albania, Bosnia, Georgia and Turkey) are ‘designated as safe countries of origin’ at EU level, save in ‘one or more’ of three circumstances:

(a)    there is a serious and individual threat to a civilian’s life or person by reason of indiscriminate violence in situations of international or internal armed conflict in the country;

(b)    restrictive measures within the meaning of Title IV of Part Five of the Treaty on the Functioning of the European Union have been adopted in view of the country’s actions;

(c)     the proportion of decisions by the determining authority granting international protection to the applicants from the country - either its nationals or former habitual residents in case of stateless persons – is higher than 20% according to the latest available yearly Union-wide average Eurostat data.

The first of these tests replicates the wording of one of the grounds for ‘subsidiary protection’ in EU law on qualification for status, although there is no cross-reference to that legislation here in this context. Among the candidate countries, the only one which might be subject to this rule is (obviously) Ukraine, as long as the Russian invasion persists. The CJEU has recently been asked whether individual applications for subsidiary protection are even possible given that those fleeing Ukraine have temporary protection; but arguably the wording of the new proposal raises a different issue, because in this context the existence of the threat would be judged as regards the situation in the country concerned more broadly, rather than in the context of an individual application for protection. Also, if the drafters had wanted an exception regarding temporary protection, they would surely have provided for it expressly; and anyway Ukraine will likely be covered by the third test.

The second test refers to EU foreign policy sanctions. A quick look at the EU sanctions database informs us that arguably none of the countries concerned face sanctions because of the country’s actions: the sanctions as regards Ukraine and Moldova relate to the actions of Russia or Kremlin surrogates; the sanctions as regards Mediterranean drilling concern only certain Turkish businesses; and the sanctions relating to Serbia and Montenegro are expressly described as historic (relating to claims as regards the previous Yugoslav war). (The recent EU sanctions against Georgia are a visa measure, not a foreign policy measure).

The third test flips the new ‘low recognition rate’ ground for accelerated procedures, meaning that neither that ground for accelerated procedures nor the ‘safe country of origin’ ground can apply once the recognition rate goes above 20%. Note that this test only takes account of first instance decision-making; if successful appeals take the recognition rate for nationals of a candidate country above 20%, that country nevertheless remains a ‘safe country of origin’ EU wide. Unlike the ‘low recognition rate’ rule as it usually applies, there is no reference to categories of people who have higher recognition rates, taking into account (for instance) appeal decisions. However, arguably significant changes’ in the country concerned must still be considered – in the context of suspending the country concerned from the common EU list, as discussed below.

Applying the third test in practice, the most recent annual Eurostat asylum statistics (2023) show a first-instance recognition rate of 2.8% for Montenegro, 6.4% for Bosnia, 1.9% for Serbia, 0.6% for North Macedonia, 7.8% for Georgia, 10.2% for Albania, 93.8% for Ukraine, 2.6% for Moldova, and 21.1% for Turkey. So on this basis, Ukraine and Turkey will not be on the EU-wide ‘safe country of origin’ list if the proposal is adopted as it stands – although the position might change on the basis of the annual asylum Eurostat statistics for 2024, which will likely be available by the time it is adopted, and the position for each candidate country may change annually after that.

Although the proposal would, in effect, create a distinct rule applicable to candidate countries as far as being ‘safe countries of origin’ is concerned, it still refers to those countries being designated as having that status. So arguably the rules for suspending that designation in the event of ‘significant changes’, and the corollary limits on Member States subsequently placing the suspended countries on their national ‘safe country of origin’ lists, continue to apply – even though these rules refer back to the general rules on designation of ‘safe countries of origin’, rather than the proposed new lex specialis rules for candidate countries (see Articles 63(1) and 64(3) of the 2024 Regulation).

The proposed specific rules for candidate countries as ‘safe countries of origin’ can be compared to the separate set of rules for EU Member States on the same point, referred to above – although the rules for EU Member States remain much more restrictive (it is far harder for nationals of EU Member States to rebut the presumption of safety, for instance; although as they enjoy free movement rights, the need to apply for international protection status to stay in another Member State will usually be immaterial for them)

The Commission’s rationale for the special rules on candidate countries is that they have already gone through a form of screening, when the European Council decided to confirm their status as candidate countries, applying the ‘Copenhagen criteria’: the ‘stability of institutions guaranteeing democracy, the rule of law, human rights and respect for and protection of minorities; a functioning market economy and the ability to cope with competitive pressure and market forces within the EU; the ability to take on the obligations of membership’. Therefore the Commission did not assess these countries against the usual criteria to be designated as ‘safe countries of origin’, as the current Article 62(1) of the 2024 Regulation would require; indeed, as noted already, the proposal would replace the current Article 62(1). However, despite the deletion of that provision, the proposed Regulation still assumes (in the preamble) that the safeguards of being a national of the supposed ‘safe country of origin’ (or a stateless person habitually resident there) and the possibility of rebutting the presumption of safety in individual cases continue to apply.     

Other countries

The seven other countries to be designated as ‘safe countries of origin’ EU wide are listed in a proposed new Annex to the 2024 Regulation. These countries are Bangladesh, Colombia, Egypt, India, Kosovo, Morocco and Tunisia. In each case, the Commission’s explanatory memorandum (and the preamble to the proposed Regulation) attempts to justify the inclusion of these countries on the list individually.

According to the Commission, the process to determine the list was based on the EU Asylum Agency, at the request of the Commission (reflecting the role of the Agency in drawing up the proposal for the common EU list, as set out in Article 62(3) of the 2024 Regulation), setting out a methodology:

to support the identification of the countries that could be considered for possible designation as “safe countries of origin” at Union level, including EU candidate countries and one potential candidate; countries of origin that create a significant asylum caseload in the EU with an EU-wide recognition rate of 5% or lower; visa-free countries that create a significant asylum caseload in the EU with an EU-wide recognition rate of 5% or lower; countries that feature in the existing Member States’ lists of “safe countries of origin” [scare quotes added]

The Commission then asked the Agency to produce country of origin information to support the Commission’s assessment; it claims that the Agency’s analysis is based on a wide range of sources:

comprising, but not limited to: European Commission reports, including the EU enlargement reports; reports by the European External Action Service; reports from the EU Agencies (such as the EU Agency for Fundamental Rights); reports from the United Nations High Commissioner for Refugees and other international organisations (e.g., the Council of Europe, the Office of the United Nations High Commissioner for Human Rights) and non-governmental organisations; political analyses from policy and international relations think-tanks; verified online media articles; newspaper articles, as well as national legislation in the countries concerned.

This can be compared to the list of sources referred to in Article 61(3) of the 2024 Regulation:

The assessment of whether a third country is a safe country of origin in accordance with this Regulation shall be based on a range of relevant and available sources of information, including information from Member States, the Asylum Agency, the European External Action Service, the United Nations High Commissioner for Refugees, and other relevant international organisations, and shall take into account where available the common analysis of the country of origin information referred to in [the Regulation setting up the Agency].

(Note that the Regulation refers to information from the Member States, but the proposal does not refer expressly to using this source for the assessment) That list of sources must be applied to establishing the EU list too, according to the current Article 62(1) of the 2024 Regulation (EU designations must be ‘in accordance with the conditions laid down in Article 61’); although, as discussed above, the Commission proposal would delete this provision.  

However, in any event it is impossible to assess either the country of origin information or the methodology developed by the Agency, because (at time of writing) the text of these documents is neither supplied by the Commission nor available on the Agency’s website. (There are some country of origin reports for some of the countries on the proposed list on the website, but those reports are outdated: 2016 for the Western Balkans; 2022 for Colombia; and May 2024 – before the demise of the previous government – for Bangladesh)  This is in spite of the Advocate-General’s opinion in the pending case of Alace and Canpelli, which  argued that Member States’ assessments underlying the designations of ‘safe countries of origin’ had to be public. (The Commission does not tell us whether any additional countries were considered for inclusion on the common list, but rejected)

Instead we have the Commission’s brief summary, starting with the assertion that ‘there is, in general, no risk of persecution or serious harm’ in these countries. This reflects part of the criteria for listing non-EU countries as ‘safe countries of origin’ set out in Article 61(1) of the 2024 Regulation (again, as noted above, the current Article 62(1) of that Regulation requires the common EU list to comply with the ‘conditions’ in Article 61; but the Commission proposes to delete the current Article 62(1)). Those criteria require that assessment to take place ‘on the basis of the legal situation, the application of the law within a democratic system and the general political circumstances’ of the countries concerned; Article 61(4) furthermore requires assessment of the application of national law, whether the country concerned complies with the ECHR or the UN’s International Covenant on Civil and Political Rights, the expulsion of citizens to unsafe countries and the existence of effective remedies to protect human rights.

The Commission then assesses each country in turn, summarising such factors as national designations, the recognition rates, human rights treaty status, national legal frameworks, democratic standards, judicial independence and impartiality, removal of citizens to unsafe countries, and the existence of persecution, the death penalty and torture. This assessment broadly reflects the criteria set out in the 2024 Regulation, although ‘recognition rates’ are not expressly referred to as part of the criteria for assessing what is a ‘safe country of origin’ (nor are trade issues or the situation of refugees from other countries, which the Commission raises in some cases).

Each of these countries gets the nod as ‘safe’ despite concerns about threats to some groups of people (such as journalists, LGBT people or women). Despite wanting to allow Member States to create exceptions to their ‘safe’ country lists for groups of people or parts of countries, the Commission does not recommend that the EU have any exceptions for any groups, even though the 2024 Regulation expressly provides that the common EU list can include such exceptions (Article 61(2) of that Regulation), and for every country on the proposed list except Kosovo, the Commission admits that there are ‘specific challenges faced by certain groups in the country which may merit particular attention’, and the preamble to the proposal states that ‘certain categories of applicants may find themselves in a specific situation in the third countries designated and may therefore have a well-founded fear of being persecuted or face a real risk of suffering serious harm’. Similarly, although the Commission notes that there are risks in particular parts of Colombia, it simply suggests that potential asylum-seekers should have moved within that country (known as the ‘internal flight alternative’), rather than propose a territorial exception to the designation of Colombia as ‘safe’.

In light of this, it is questionable why there are no exceptions for groups of people or parts of a country, particularly when the same proposal claims that, for Member States, such exceptions ‘offer means of managing likely unfounded applications efficiently while maintaining necessary legal safeguards’ (my emphasis). It seems that sauce for the Member State goose is not sauce for the EU gander; and in fact, it is arguable that the assessment of the ‘safety’ of the countries concerned is inadequate because it did not consider whether such exceptions should be granted. Of course, human rights NGOs may well have further critiques of the details of the Commission’s brief assessments of ‘safety’.

Conclusions

The new proposal is cynical in many respects. First of all, the Commission wants some restrictive rules from the 2024 Regulation to apply in advance, but not some of the safeguards that apply to them – a form of ‘cherry-picking’. This is particularly relevant to the early use of the ‘low recognition rate’ rule in the context of border procedures, without the safeguards applicable to border procedures in the 2024 Regulation, especially the exemption for unaccompanied minors. In fact, as we have seen, the Commission does not even mention or justify its proposed advance application of these rules to unaccompanied minors – still less its attempt to waive an exception that would otherwise apply to them, even when it means they can be detained.

Secondly, the Commission wants to drop the requirement to apply the usual conditions that apply to designation of ‘safe countries of origin’, not only for candidate countries (which will be subject to special rules of their own) but in general. It is possible that this is simply down to poor legislative drafting, as despite the proposed abolition of the current Article 62(1) of the 2024 Regulation, the preamble to the new proposal assumes that key safeguards continue to apply in the context of the EU common list; and the rules on suspension of designation and the corollary limits on national designation of ‘safe countries of origin’, which refer back to the general rules on designation of ‘safe countries of origin’, expressly continue to apply.

So although it is obviously questionable in principle both to drop the requirement that the common EU list is subject to the same conditions as national lists (a blatant double standard), without even replacing it, and to create a separate rule for candidate countries, the overall impact of this change is blunted. Nevertheless, it would be better in principle to retain a single common standard for designation of ‘safe countries of origin’; it is particularly objectionable to have double standards compared to national lists and even more so, no standards at all for the EU list.

The Commission’s lack of transparency of its sources for assessing the group of countries to go on the common list is likewise questionable; and its treatment of the candidate countries is simply opaque. The countries concerned are not even named, and the Commission offers no interpretation of its proposed new criteria relating to these countries, or a discussion of how they would apply in practice. (The application of the ‘low recognition rate’ rules is also opaque, in the absence of a simple list of the recognition rates by country)

All in all, this proposal is both murky and unprincipled: an unimpressive start to the next phase of EU asylum law.  

 

 

 

 

 

Saturday, 19 April 2025

Hamoudi v Frontex: Advocate General Norkus’ Opinion - Reversing the Burden of Proof and the Presumption of Frontex’s Privileged Access to Evidence


 

Antje Kunst*

*Antje Kunst is an international lawyer and barrister of Garden Court North Chambers, admitted to the Bar of England and Wales, and the Bar of Berlin, advising and representing individuals in a wide range of matters related to fundamental rights within the CFSP and other fields. She has appeared in numerous cases before both the Court of Justice and the General Court, within the Court of Justice of the European Union.  

Photo credit: Francesco Placco, via Wikimedia Commons

 

Introduction

Following the hearing  on 4 February 2025 by the Grand Chamber on the appeal of Hamoudi v Frontex in Case C-136/24 against the General Court (GC)’s Order of 13 December 2023 the Advocate-General rendered his opinion on 10 April 2025 (‘Opinion’).  

As outlined  here, the case of Hamoudi v. Frontex involves a Syrian asylum seeker, Mr. Alaa Hamoudi, who, claims that on 28 and 29 April 2020 he was a victim of collective expulsion in the Aegean Sea. In this context, Mr. Hamoudi asserts that while a group of 22 individuals including himself were at sea, a private surveillance aircraft, equipped with a camera and operated by the European Border and Coast Guard Agency (Frontex), flew over the scene twice. A highly damaging report by the EU Anti-Fraud Office (‘OLAF report’) on Frontex was made public through an NGO in October 2022 and covered precisely the period in question — including April 2020 — and confirmed that incidents of pushbacks occurred under Frontex's watch, involving Frontex surveillance aircraft in the Aegean Sea.

In an action for damages before the EU General Court, Mr. Hamoudi requested that he be compensated by Frontex for the non-material damage suffered due to the collective expulsion. In his arguments before the General Court Mr. Hamoudi inter alia asserted that because Frontex surveillance aircraft flew over the collective expulsion it had specific knowledge of that incident but failed to report it and take appropriate action as per its mandate (see para. 11 of the Opinion)

The General Court dismissed the action as manifestly lacking any foundation in law. It found that Mr. Hamoudi had failed to prove the actual damage he claimed to have suffered. The evidence adduced by the appellant ‘was manifestly insufficient to demonstrate conclusively that he was present at and involved in the alleged incident of 28 and 29 April 2020’ (see paras. 39 and 62 of the Court Order).  

Advocate General's limited Analysis: Reversal of the Burden of Proof

As requested by the Court of Justice of the European Union (‘Court of Justice’ or ‘CJEU’) the Advocate General centred his analysis on a possible reversal of the burden of proof in relation to the existence of damage in collective expulsion cases. (para.22 of the Opinion).  In doing so, he looked in detail at the case-law of the Court of Justice in various areas including discrimination (paras. 30 to 37 of the Opinion) and the case law of the European Court of Human Rights (ECHR) in expulsion cases (paras. 40 to 50).

In his analysis of the case-law of the ECHR in expulsion cases he considered the recent successful collective expulsion case of A.R.E. v Greece in particular ‘instructive’.

Importantly, the Advocate General referred to the fact that in A.R.E., the ECtHR applied its case law on the burden of proof related to secret detention cases. Once the applicant presents prima facie evidence, the Court may draw strong adverse inferences from the respondent government’s failure to disclose essential documents to establish the facts, or failure to offer a satisfactory and convincing explanation of the alleged events. (paras. 41 to 44 of the Opinion)

In line with the jurisprudence of the CJEU and ECtHR, the Advocate General rightly emphasizes that the burden of proof cannot be placed on the claimant to establish facts or produce evidence that is likely to be exclusively in the possession or control of the respondent, here Frontex. If any concrete evidence of collective expulsions exists, it is far more likely to be held by the alleged perpetrator rather than the victim. (para. 51 of the Opinion)

The Advocate General’s three conditions

In the Advocate General’s view, based on the CJEU’s and ECtHR’s case law, the burden of proof may be reversed in cases like Mr. Hamoudi’s case when three conditions are met.

The FIRST CONDITION: (see para. 57 of the Opinion) is that the claimant must present prima facie evidence in support of his or her claim. If the claimant’s account is inconsistent, incoherent, or if the claimant lacks credibility, this initial burden is not met, and the case should be dismissed. Consequently, the reversal of the burden of proof becomes relevant only once a prima facie case has been established. In this context, the Advocate General relies in particular on the case law of the ECtHR, for example as set out in the recent case of G.R.J. v Greece (see § 179).

As stated here, from the questions of the judges at the hearing it appears that the Court of Justice is considering to hold that prima facie evidence had been presented by Mr. Hamoudi. In any event, the Advocate General correctly opines that the General Court may have placed the ‘evidentiary bar’ too high in the first place (para. 64 of the Opinion). In this context, though, the Advocate General omits that the General Court in its assessment of the evidence failed to consider the various contradictory public statements about the events by Frontex outside the proceedings, in particular when the OLAF report was made public and crucial general contextual evidence. Such general contextual evidence was considered in the recent G.R.J. and A.R.E. cases by the ECtHR. This evidence led the ECtHR to conclude that there is a systemic practice of pushbacks in the Aegean See and that the Greek Government failed to refute the evidence by providing a satisfactory and convincing alternative explanation (see § 190 in G.R.J. and § 229 in A.R.E. and footnote 80 of the Opinion).

The SECOND CONDITION: (para. 59 of the Opinion) for the burden of proof to shift is the existence of a clear or structural imbalance in access to evidence—specifically, where the claimant faces significant obstacles in presenting evidence, while the respondent, in this case Frontex, is in a better or more privileged position to refute the allegations. Mr. Hamoudi, a vulnerable asylum seeker with his mobile phone confiscated, lacks the evidence to prove Frontex’s involvement in the events, whereas the respondent is best placed, in a privileged position to prove or disprove the allegations (see in more detail under: ‘Presumption of privileged access must be applicable to Frontex’). There can be no doubt that the second condition is met.

The THIRD CONDITION: (para. 60 of the Opinion) is that the failure to shift the burden of proof would render ineffective the claimant's (fundamental) rights protected under EU law while a shift would not undermine the respondent’s (fundamental) rights under EU law. In Mr. Hamoudi’s case the failure to shift the burden of proof would render ineffective his fundamental rights inter alia under Article 19 of the EU Charter of Fundamental Rights and undermine his right to an effective remedy under Article 47 of the Charter. The shift would not undermine any (fundamental) right of Frontex under EU law. On the contrary, by providing evidence to prove or disprove the allegation, Frontex assists the Court in reaching an accurate outcome (regarding a public authority’s duty pursuant to Article 24 of the Court’s Statute to assist the Court in reaching the correct result see more here).

Presumption of privileged access to evidence not applicable?

Somewhat surprisingly towards the end of the Advocate General’s Opinion, he expresses the view that the presumption established in the jurisprudence of the ECtHR on expulsion cases—namely, that the claimant is at a disadvantage in presenting evidence while the respondent is in a stronger or more privileged position to rebut the allegations – may not be applicable to Frontex. The ‘automatic’ shift of the burden of proof, once prima facie evidence has been presented, allegedly could only be applied when the authorities of a Member State are involved (paras. 61 and 62 of the Opinion). This despite the fact that the AG opined earlier in his Opinion (para. 29) that the differences in the procedures before Strasbourg and the CJEU are more apparent than they are real.

According to the Advocate General, actors like Frontex possess more limited powers compared to the authorities of a Member State, and it was unclear whether—and to what extent—their actions contribute to the difficulties claimants face in adducing evidence of their involvement in the events at issue. Due to these limited powers, it was not clear that they would be in a better or more privileged position to rebut the claimant’s allegations (para. 62 of the Opinion).

The Advocate General’s suggestion that Frontex’s actions must have contributed to the difficulties experienced by the claimants in adducing evidence of their involvement in the events at issue is misplaced. This is not a requirement under the case law of the Court of Justice, nor under that of the ECtHR. According to this case law, it is sufficient to establish the evidentiary difficulties faced by the applicant and the respondent’s capacity to provide evidence in rebuttal.

It is wholly irrelevant whether a Member State may have more authority over certain types of evidence.  The issue is not whether Frontex differs from a Member State in terms of powers over evidence but rather who holds relevant evidence, and who does not. This must be assessed in casu,  e.g. here in relation to Hamoudi and Frontex. The Agency is not being asked to produce evidence concerning Member States’ actions, but rather evidence regarding its own actions or inactions, and knowledge, particularly in light of its established presence and two active joint operations in the region.

At the hearing, Frontex claimed its hands were tied because it did not have access to evidence held by the Member State. This is irrelevant, what matters is that it has evidence in its possession and control that may reveal what the agency did, failed to do, or knew regarding the alleged collective pushback action on 28 and 29 April 2020 in the Aegean Sea.

Undesirable Consequences of having to demonstrate Privileged Access to Evidence

If, as the Advocate General proposes, the burden of proof only shifts when it is first shown that Frontex is better placed to refute the allegations than a claimant, then in practice, the burden might never shift. The Court of Justice should firmly reject this approach.

Accepting that Frontex is - unlike a Member State - not presumed to have privileged access to evidence would undermine its positive obligations to protect fundamental rights of individuals in distress at sea and enable Frontex’s impunity and outright ignores a textual reading of article 7(4) in line with article 80(2) and 80(3) of the Frontex Regulation, which outlines the exclusive positive human rights responsibilities that Frontex has within the context of joint operations. It would risk making Frontex’s legal obligations unenforceable. This would also contradict the Advocate General’s THIRD CONDITION outlined in paragraph 60 of the Opinion.

Presumption of privileged access must be applicable to Frontex

Related to Frontex’s own actions, its own responsibility to comply with its own fundamental rights obligations as per the Frontex Regulation, and the harm resulting for claimants, Frontex must be presumed of being in a better or more privileged position in collective expulsion cases such as the present one.

The appellant is a Syrian refugee, pushed back at night, with his mobile phone confiscated and lacks access to the evidence which shows Frontex’s involvement. Frontex, on the other hand, operates its own aerial surveillance systems and joint operations logs and collects video and radar data from its flights. The likelihood that it possesses or has under its control relevant video surveillance footage and/or incidents reports, relevant logbooks on surveillance operations is extremely high which places Frontex in a particularly privileged position with regard to access to evidence in these types of cases. Moreover, it was established in the OLAF report and during the hearing that Frontex was fully aware of the practice of so-called ghost landings pursued by the Hellenic coast guard and from that flows the duty within the context of its joint operations to collaborate with Member State authorities to prevent human rights abuses.

While it is true, as the Advocate General points out, that the General Court did not examine Frontex’s involvement in or knowledge of the alleged events of 28 and 29 April 2020, this omission stems from the General Court’s erroneous exclusive focus on the question of damage to the appellant and whether he was present and affected by those events. However, this cannot mean that Frontex cannot be regarded as possessing or having under control the evidence needed to rebut Mr. Hamoudi’s allegations. (c.f. para. 63 of the Opinion). The very nature of ghost landings entails that the Hellenic Coastguard is no longer present at sea.

Accordingly, Frontex is in the exclusive position and best placed to confirm or deny the prima facie evidence provided by Mr. Hamoudi regarding the events that took place at sea, including whether it had a surveillance aircraft with a camera operating over the area during the collective expulsion of 28 and 29 April 2020 but failed to report it.

If there is an shift of the burden of proof based on the available contextual and personal prima facie evidence, Frontex would not face a probatio diabolica (an impossible proof). It is also not an unreasonable proof for Frontex to provide: recall that its access  led to the damning  report by OLAF which affirmed its presence on the night of 28-29 April. Frontex is not asked to prove facts that lie completely outside its sphere of influence and knowledge (see the case law of the CJEU cited at para. 53 of the Opinion). To the contrary: its being asked to provide evidence regarding events taking place within the very area of the joint operations in the region, falling entirely and exclusively within its mandate.

Conclusion

The Court of Justice in the present case should accept that the three conditions proposed by the Advocate General must be met in expulsion cases like this one for the burden of proof to shift. It should hold that this entails a reversal of the burden of proof for Frontex when the case concerns its own actions or inactions.

Contrary to the Advocate General’s assertion in his Conclusion, the state of the proceedings before the General Court permits the Court of Justice to assess that the appellant adduced prima facie evidence (FIRST CONDITION) which does allow for the burden of proof to shift, as Frontex is in a better and more privileged position than the applicant to prove or disprove its involvement in, and knowledge of, the alleged events (see Footnote 96 of the Opinion). This is not a question for the General Court to determine in proceedings following a referral back to it (contrary to what the Advocate General proposed in his Conclusion of his Opinion).

It is well established that Frontex possesses evidence directly linked to its core (and exclusive) responsibilities, including the gathering of surveillance data, the production of incident and operational reports, and the monitoring of compliance with fundamental rights as per Article 7(4) and 80(3) and 80(4) of the Frontex Regulation. This places the agency in a similarly ‘privileged’ evidentiary position as Member States are in relation to their own responsibilities in this case.

The Court of Justice should therefore set aside the order under appeal, hold that the three conditions for shifting the burden of proof as outlined by the Advocate General are met in Hamoudi’s case as it would have been the case if the respondent had been a Member State (see footnote 96 of the Advocate General’s Opinion). It should refer the case back to the General Court to reassess the matter, taking into account that the burden of proof has shifted to Frontex.

Between pragmatic and legal considerations: comment on the Advocate General's opinion in joined cases C‑758/24 [Alace] and C‑759/24 [Canpelli]

 

 


 

 

Matteo Zamboni (human rights lawyer working between Italy and the UK. He is a partner to the Immigration Law and Policy clinic of Goldsmiths, University of London)

 

Photo credit: Matteo Zamboni - The Grand Chamber of the CJEU in Luxembourg before the hearing of 25 February 2025

  

Introduction

 

On 10 April 2025, Jean Richard de la Tour, Advocate General (AG) at the Court of Justice of the European Union (CJEU), delivered his conclusions in joined cases C‑758/24 [Alace] and C‑759/24 [Canpelli], regarding the implementation of the Italy-Albania Protocol on asylum and return procedures, which was the subject of my previous blog.

 

In a nutshell, as stated in the press release from the Court, the AG argued that (i) ‘a Member State may designate safe countries of origin by a legislative act’; (ii) a Member State ‘must disclose, for the purpose of judicial review, the sources of information upon which that designation is based’; (iii) ‘[a] Member State may […] under certain conditions, grant a third country the status of safe country of origin, while identifying limited categories of persons likely to be at risk of persecution or serious harm in that country’.

 

This post tries to dig a little bit deeper in the opinion in order to highlight its nuances, strengths, and criticalities.

 

Context

 

Some background information helps put the opinion into context.

 

The opinion was published on the same day that the Italian military vessel Libra (already infamous for not taking action when called to the rescue of the victims of the massive shipwreck of 11 October 2013) was sailing toward the centres in Albania to bring 40 individuals whose claims for international protection had been rejected by the Italian authorities. According to the information available, these people ‘will be held in [the] Italian-run detention centres until they are repatriated to their home countries’. In fact, it was reported that, ‘facing criticism from the opposition over the legal confusion, Prime Minister Giorgia Meloni's conservative coalition […] decided to use [one of the two detention centres built in Albania, the one in Gjader] as a staging post for people whose asylum bids have already been turned down’. More precisely, by means of decree-law no 37 of 28 March 2025 the Italian government established that one of the two centres built in Albania  will temporarily be turned into a repatriation hub – in Italian, ‘Centri di permanenza per i rimpatri’, abbreviated CPR; that is to say, centres to detain undocumented migrants whose claims to international protection had been rejected with a final decision and who are awaiting deportation to their countries of origin.

 

Interestingly, the amendment enacted by the Italian government mirrors the proposal of the EU Commission for a Common European System for Returns, which, amid criticism from human rights organisations, ‘endorsed the idea of “return hubs” located outside the European Union’. Yet, according to sources in the press, the plan is only temporary. In fact, ‘the government still hopes to revert to its original plan and is waiting on a ruling from the European Union's Court of Justice which could compel Italian judges to process new asylum seekers dispatched to Albania’.

 

Moreover, on 16 April 2025 the EU Commission published a proposal for amendments to regulation 2024/1348 (the asylum procedures Regulation, forming part of the asylum pact) touching precisely on the matter of the designation of safe countries of origin (SCOs). Notably, the Commission highlighted that ‘Regulation 2024/1348 […] for the first time provides for the possibility to designate safe countries of origin at Union level’ and informed that ‘the EU Agency for Asylum (EUAA) had been asked to accelerate its analysis […] with a view to drawing up an EU list [of SOCs]’.

 

The proposal is of interest for the case under review for at least three main reasons: 

 

(i)                from a general perspective, the European Commission made it clear that the proposed amendments have an eminently practical purpose: ‘[to] help Member States manage asylum applications more efficiently’ and with ‘greater flexibility’;   

(ii)              the Commission took note of the fact that Article 61(2) of regulation 2024/1348 ‘allows for designation of […] safe countries of origin with exceptions’, including ‘by excluding specific regions or clearly identifiable categories of individuals’, and proposed to ‘bring[…] forward [the] application [of these exceptions]’; the Alace and Canpelli case is partly about the designation of categories of asylum seekers who may be at risk even though their country of origin is designated as ‘safe’;

(iii)           the provisional list of SCOs proposed to be designated at EU level includes Bangladesh; i.e., the country of origin of Alace and Canpelli, the fictitious names of the asylum seekers concerned by the present case – although the Commission does not propose any exceptions for categories of people in that country.

 

Summary of the AG opinion

 

The context briefly discussed above, to which it may be added the fact that pictures of the 40 asylum seekers handcuffed at the time of disembarkation in Albania made the front pages on Italian newspapers, could explain, at least in part, the reason why the AG opinion is full of reference to non-legal considerations.

 

Indeed, the opinion cites, for 4 times in total, the ‘high migratory pressure’ against some Member States as one of the factors to take into due consideration in the effort to advance a balanced interpretation of the relevant provisions of directive 2013/32 (the current asylum procedures Directive, which applies until the 2024 Regulation takes effect).

 

From a legal point of view, the balancing exercise transpires from the juxtaposition of two concepts: one the one hand, that of the margin of discretion (and/or appreciation) enjoyed by Member States when giving effect to the directive at hand, cited 10 times in total; on the other hand, the doctrines of “effet utile” and of the primacy of EU law, referred to in the opinion 6 times in total.

 

Generally speaking, the careful balancing of these competing legal concepts, coupled by practical considerations regarding the State response to mass migration, resulted in the nuanced replies given by the AG.

 

The first question

 

To begin with, the AG noted that the first question, asking whether EU law precludes a national legislature from designating a third country as a SCO by a legislative act of primary law, ‘does not raise, in itself, any particular problem’. The AG noted that directive 2013/32 does neither specify which national authorities are competent for such designation, nor requires that the designation is made with a specific instrument. To the contrary, the expression ‘legislation’, contained in Article 37(1) of the directive, ‘must be understood in its broadest sense, as including acts of a legislative, regulatory or administrative nature’. As a consequence, the AG acknowledged that, under the principle of ‘institutional and procedural autonomy’, Member States enjoy ‘a wide margin of discretion’ as regards the means and the procedures to be used to proceed to the designation of SCOs. Such discretion encompasses the possibility to proceed to the designation by means of primary legislation; i.e., acts of Parliament. At the same time, the AG clarified that ‘the act by which a Member State designates third countries as [SCOs] must not affect […] the basic principles and fundamental guarantees set out in […] directive 2013/32, and in particular […] the right to an effective judicial remedy’ (see AG opinion § 35-39, all translations from the original in Italian and French into English are mine).

 

Thus, the AG replied to the first question stating that:

 

‘Articles 36 and 37 of directive 2013/32 must be interpreted as not precluding Member States from designating SCOs by means of legislative acts’. However, this is possible only insofar as (i) the primacy of EU law is guaranteed; and (ii) the obligations and the objectives of the directive are fully implemented’ (see § 39).

 

The second and the third question

 

A similar tension between the doctrine of the margin of discretion and the overarching objective to ensure the full and consistent application of the acquis communautaire throughout the European Union legal space permeates the answer given to the second and the third question, regarding the need to make publicly available the sources used to justify the designation of a country as a SCO (second question) and the possibility, for the national judge, to assess the designation by making use of information drawn independently from the sources referred to in Article 37 and Annex I of directive 2013/32 (third question). (The Annex defines the criteria to designate a country as a ‘safe country of origin’)

 

At the outset, with a clarification that was much needed after some confusion in the arguments made by the parties (and, above all, by the Italian government and the intervening Member States) during the hearing of 25 February, the AG stated that ‘these questions do not concern the review to be carried out by [the] judicial authority with regard to Article 36(1) of directive 2013/32, which requires the rebuttal of the presumption of safety of a given country [designated as a SCO] in a particular case where, as a result of a specific and detailed assessment […], it appears that that country is not safe due to the individual circumstances of an applicant’ but rather the different issue of challenging the general designation of a country as a SCO (see § 44).

That said, the answers of the AG follow the reasoning applied with regard to the first question. Notably, the AG reiterated that the designation of SCOs by means of primary legislation cannot be construed as excluding the possibility to subject the designation to proper judicial review, as required by Article 46 of the directive. Indeed, when they proceed to the designation of SCOs under Article 37 of the directive, Member States are implementing EU law. Thus, quite regardless from the means employed, they must ‘ensure the respect of the substantive and procedural guarantees established […] by EU law’. It is, therefore, ‘essential’ that ‘sufficient and adequate publicity’ is given to the sources used by the national authorities to proceed to the designation of a county as a SCO (see § 48-49).

 

The AG acknowledged that the publication of the sources is not expressly required by the letter of directive 2013/32. However, he argued that such obligation can be inferred from a ‘systemic reading’ of the directive that takes into due account its overall ‘objectives’. Indeed, ‘the rebuttable nature of the presumption that a country is safe’ implies that  the asylum seekers concerned as well as the national (judicial or administrative) authorities are allowed to know the grounds relied on by Member States to proceed to the designation of a country as a SCO with a view to guaranteeing their right to, respectively, challenge and assess the legitimacy of such designation (see § 50-51 and 54-55).

 

The conclusion is confirmed in the light of the general principle of the duty of cooperation (set out, with specific regard to refugee law, in Article 4(1) directive 2011/95) and the right to an effective remedy (guaranteed by Article 47 of the Charter), which impose that asylum seekers and national courts are given access to all relevant materials, including those allowing them to assess the possible ‘violation of the substantive conditions of [the] designation [of a country as a SCO]’ (see § 56-59).

 

Yet, such a strong petition of principle was not translated into an equally forceful practical answer. As a matter of fact, in answering the third question, the AG gave different options to Member States, arguing that they should include the sources as annexes to the legislative act adopted to designate SCOs, or, alternatively, communicate them at the request of the asylum seeker concerned and/or of the national administrative or judicial authorities. Moreover, in answering the fourth question, the AG held that ‘if the sources of information are not disclosed, then the competent judicial authority may review the legality of [the] designation [of a country as a SCO] in the light of the conditions set out in Annex I to the directive on the basis of the sources that were gathered by the judicial authorities among those listed in Article 37(3) of the directive’ (see § 62, 64, 65).

 

Admittedly, these answers may be said to fail to curb all the uncertainties, as they leave several options open to Member States who are unwilling to publish the sources on the basis of which they proceeded to the designation of SCOs. Moreover, the answer to the fourth question (and in particular the opening caveat ‘if the sources of information are not disclosed’) is capable to undermine the answer to the third question (which states that, anyway, the sources must be published).

 

Be this as it may, this does not seem problematic in the specific case of Alace and Canpelli, as the Italian government had already back-trailed on their position and, with a decision of 28 March 2025, had anticipated the judgment of the CJEU by providing for the publication of the country reports relied on for the purpose of the designation of SCOs.

 

The fourth question

 

The last question, regarding the possibility to designate a country as a SCO notwithstanding the existence of categories of persons for whom it does not meet the substantive conditions for such a designation, was rightly identified as the most complicated one, and has indeed prompted the AG to draft a very nuanced reply.

 

Namely, the AG came up with two possible solutions. The first, based on a ‘restrictive’ interpretation of Annex I to directive 2013/32, relied on the assumption that, to be considered as such under EU law, a safe country must guarantee ‘all nationals and stateless persons living there sufficient protection against the risk of persecution or serious harm, regardless of the portion of the territory in which they find themselves and [independent from] their race, nationality, political [opinions] or religious beliefs, or […] their belonging to a specific social group’. On the other hand, the second solution is meant to ‘allow [Member States] to designate a third country as a SCO even though one or more categories […] of individuals at risk […] have been identified’ (see § 68 and 70).

 

Between the two, the AG concluded that the second option would seem the correct one.

 

The reasons adduced to discard the first option seem more practical than legal. Indeed, the AG borrowed the expression used by the Italian government in their oral arguments and stated that such a restrictive interpretation of the concept of SCO under EU law would be ‘idealistic’ and would undermine the practical effect of directive 2013/32 insofar as it would prompt ‘Member States to treat all the requests advanced by citizens [coming] from those countries under the ordinary procedure, even though the overwhelming majority of those asylum seekers do not have any real need of international protection’. According to the AG, this would result in a ‘procedural congestion’ all the more intolerable in ‘a context characterised by a strong migratory pressure’ (see § 70).

 

In essence, this seems to be the reason prompting the AG to favour the second solution, albeit, as the AG himself did not fail to point out, this may be seen as being at odds with the findings of the CJEU in case C-406/22 (last year’s judgment in which the CJEU interpreted the currently applicable Directive to mean that Member States could not designate part of a country of origin as ‘safe’).

 

To overcome the contradiction, the opinion argued that the second option is grounded in (i) the letter of Annex I to directive 2013/32; (ii) the systematic reading of the relevant EU legislation; and (iii) its purpose.

 

Before addressing this three-pronged argument, however, the AG resorted, once again, to ‘pragmatic’ considerations, stating that ‘in the event that the asylum system of a Member State is under strong migratory pressure and [faces] a high proportion of manifestly ill-founded applications lodged by nationals from [SCOs]’, the second option ‘would [represent] a balanced solution which would make it possible, on the one hand, to […] expedite the examination of those applications [and, on the other hand,] to ensure [that] all applications [receive] appropriate treatment in accordance with the provisions of directive 2013/32’ (see § 71-72).

 

Having said that, the opinion focusses on the letter of Annex I, and specifically on the interpretation of the adverb ‘generally’ (included in the formula ‘a country is considered as a safe country of origin where […] it can be shown that there is generally and consistently no persecution […], no torture or inhuman or degrading treatment or punishment and no threat by reason of indiscriminate violence in situations of international or internal armed conflict’). According to the AG, the adverb ‘generally’ refers to ‘an event, a fact or any other circumstance which manifest itself in the majority of cases or […] which is applicable to a really wide number of persons, without taking into account particular cases’. It, then, follows that, ‘from a literal point of view, a third country may be designated as a SCO if it is demonstrated […] that it protects, not each of its nationals but, the majority of them’. In this scenario, as it had been argued by the Italian government, the second option would also be confirmed by the text of Whereas 42 of directive 2013/32 (see § 78-79).

 

This argument is reinforced by a systematic argument regarding the dichotomy between general and specific considerations underpinning directive 2013/32. In the opinion of the AG,

 

‘if [EU law] requires the competent national authorities to depart (ex post) from the presumption of the safety of a country whenever they determine, as a result of an individual examination […], that the person concerned may, by reason of his individual circumstances, be exposed to a risk of persecution or serious harm in his or her country of origin, then [there is] no valid reason why a Member State should not decide, as a result of the general assessment of that country, to exclude (ex ante) from the scope of that presumption the category or categories of persons whom it has […] identified as being at risk’ (see § 81).

 

Finally, the second solution is said to fit the purpose of the directive, which is to allow Member States to ‘speed up the procedure […] every time that a request for international protection may be ill-founded’. From a wider perspective, this is also confirmed by the consideration of the ‘margin of appreciation’ enjoyed by EU Member States when enacting the directive. Moreover, in reaching this conclusion the opinion also relies on the new provisions of regulation 2024/1348, which, at Article 61(2), explicitly allows for the possibility to designate a country as a SCO even though there exist categories of individuals at-risk. As a matter of fact, and even though the new regulation will become applicable only as of 12 June 2026 (unless the Commission’s recent proposal to bring forward parts of it is adopted beforehand), the AG noted that it would be ‘paradoxical’ to prevent Member States from using this possibility at a moment in which they are called to ‘adequately prepare to implement’ the new regulation, including Article 61(2) (see § 83, 85, 94). 

 

At the same time, the AG seemed to be conscious of possible abuses, and indeed strived to place clear limitations to the implementation of the second option by requiring that these categories are ‘limited’ and ‘clearly identifiable’. In general terms, the need for a qualified application of the personal exceptions to the designation of SCOs is predicated on the assumption that the margin of appreciation, or discretion, of Member States is limited by EU law and by the principle of proportionality. As a result, the use of discretion can never ‘impair the general objectives of directive 2013/32’ (see § 85-87 and 93). 

 

As a consequence, Member States must ‘confine [such] personal exceptions to a very limited number of persons’. Otherwise, the very operation of the presumption of safety would be questionable. In other words, if a given country is designated as a SCO notwithstanding the identification of numerous categories of people that may be exposed to the real risk of persecution of serious harm (as for example, all members of the LGBTQIA+ community), then the concept of safe country of origin would be tantamount to a ‘legal fiction’ (see § 70 and 91-92).

 

Based on this (quite balanced and nuanced) reasoning, the opinion concluded that

 

‘Articles 36 and 37(1) of, and Annex I to, Directive 2013/32 must be interpreted as not precluding a Member State from designating a third country as a safe country of origin for the purposes of examining applications for international protection, while at the same time identifying limited categories of persons as potentially exposed to a risk of persecution or serious harm in that country, provided that on the one hand, that the legal and political situation in that country characterises a democratic regime under which the population enjoys, in general, durable protection against that risk and, on the other, that Member State proceeds accordingly to expressly exclude those categories of persons from the application of the concept of safe country of origin and the presumption of safety attached to it’ (see § 95).

 

Final considerations

 

Few conclusive remarks further highlight the very nuanced nature of the AG opinion.

 

This is particularly evident as regards the first three questions, the answers to which really give the impression of a balancing exercise between the position of the Italian government (and the intervening Member States) and that of the defence. Indeed, the AG validated the practice to designate SCOs by means of primary legislation while at the same time making it very clear that this practice cannot be invoked in order to undermine the guarantees set out at EU level.

 

All in all, the AG found a workable compromise between the legal concepts of the margin of discretion/appreciation and the fundamental principle of the consistent application of the acquis communautaire and its supremacy over national law.

 

On the facts of the case, this approach resulted in a very careful stance toward what seems to be the crux of the questions. Indeed, the AG does not mention the possibility to file a challenge of constitutionality with the Italian constitutional court, which was forcefully put forward by the Italian government as the effective remedy provided by the domestic legal system to question the designation of a country as a SCO. In so doing, the opinion confirmed that the judicial review over the legislation enacting the provisions of directive 2013/32 must be performed by the judge of the case, including first-instance judges. At the same time, the AG opinion does not go as far as stating that the national judge is empowered to disapply national legislation in case of non-compliance with the provisions of the directive. Rather, the multiple references to the doctrine of “effet utile” seem to favour the interpretation that, as I had already argued, the case at hand concerns provisions of EU law lacking direct effect.

 

Still, the answers to the first three questions seem to find a reasonable balance between legal and non-legal considerations. Unfortunately, it seems that one could argue that the matter is different with regard to the answer to the fourth question, as it appears that the option to favour the solution allowing for personal exceptions to the designation of SCOs is grounded on ‘pragmatic’ considerations more than on sound legal reasoning.

 

In particular, this seems to be the case as regards the literal argument employed by the AG, which focusses on the term ‘generally’, but completely overlooks the other adverb in the hendiadys; that is, in the English version of the directive, ‘consistently’. According to the English dictionary, consistently means ‘in every case or on every occasion; invariably’. Therefore, it seems that the literal meaning of this term would rather support a more restrictive interpretation of Annex I; i.e., the interpretation excluding the possibility to designate as SCO a country which is not safe for specific categories of persons.

 

True, the opinion justified this approach by making reference to some inconsistencies in the different language versions of Annex I to directive 2013/32. However, the language versions that I have examined (i.e., the English version, that says ‘generally and consistently’; the French version, which states ‘d’une manière générale et uniformément’; and the Italian version, in which these terms are translated as ‘generalmente e costantemente’) seem to have the same meaning.

 

Also the reference to recital 42 in the preamble to directive 2013/32 may be said to cause raised eyebrows. In fact, when taking into due account the distinction (highlighted by the AG at § 44 of his opinion) between a (more specific) challenge to the application of the concept of SCO to a given case, and a (more general) challenge to the designation in itself, it seems that recital 42 refers to the latter question, while the matter addressed in the fourth question pertains to the second.

 

Finally, the argument regarding the future application of Article 61(2) of regulation 2024/1348 appears to be contradicted by the latest proposal from the EU Commission. Indeed, when the Commission proposes to ‘bring forward’ the application of the personal and territorial exceptions provided in the new legislation, it seems to imply that the EU legislative framework as it currently stands does not allow for such exceptions.

 

Hopefully, these aspects will be clarified by the CJEU in the judgment, which is expected before the summer.