Italy's Albania experiment and the legal collapse of 'safe third country'

On 16 October 2024 an Italian coastguard vessel delivered 12 male migrants from Bangladesh and Egypt to the port of Shengjin, Albania. They were the first arrivals under the Italy–Albania Protocol, the five-year agreement signed by Prime Minister Giorgia Meloni and Albanian Prime Minister Edi Rama that allows Rome to process asylum claims for adult male migrants intercepted in international waters at purpose-built centres on Albanian soil. Within weeks the cohort had been returned to Italy after Italian courts ruled that the countries of origin named in their applications were not, as a matter of law, 'safe.' Another group of seven arrived in February 2026 and was likewise repatriated by judicial order. The legal scaffolding around the policy is the focus of recently completed research by Dr Gaia Romeo, whose paper, published in mid-2026 in the Journal of International Migration and Integration (Springer / ScienceDirect), examines the only instance in which the 'safe third country' concept has been applied to large numbers of asylum seekers inside the European Union, and the only one in which it has produced a track record of court-ordered reversals.

Romeo's central finding is that 'safe third country,' as it currently operates in EU and member-state practice, is less a legal standard than a procedural alibi. The concept is invoked to relocate the place where an asylum claim is examined, not to guarantee the substance of that examination. In Italy's Albanian protocol, the design assumes Bangladesh, Egypt, Tunisia and a rotating list of others can be designated 'safe,' which shortens the interview at first reception and permits accelerated removal. The premise is that the designated third country offers effective protection in cases where the country of origin does not. The empirical record produced by 2026 shows the premise does not survive contact with either set of courts, Italian or European.

The legal shell

The procedural mechanics are the story. The Italian protocol's expedited-track logic follows the EU Asylum Procedures Directive, which permits an application to be declared inadmissible when the applicant could have found protection in a third country on the journey in. The 'safe third country' doctrine dates back to a 1996 Council of the European Union resolution, was carried into the Procedures Directive (2013/32/EU), and has since been debated in academic and policy circles, including at the European Policy Centre. Romeo's contribution is to subject the doctrine to a single-application test: does the instrument, in practice, grant substantive rights to asylum seekers, or only procedural displacement.

The answer, on her evidence, is the latter. Italian judges found that the first cohort's return to Bangladesh or Egypt would expose them to persecution or harm that European asylum standards consider material. The reasoning pointed to specific country-of-origin information from UN agencies. The 'safe' designation, in other words, was not a finding about conditions on the ground. It was a procedural ticket whose validity was not tested against those conditions.

The chain of designation

What makes the Albanian protocol unusually informative for EU policy is the chain of legal moves behind it. The Italian government first negotiates a bilateral readmission agreement with a transit country (Albania, in this case). It then relies on the EU-level 'safe third country' concept to process applications without substantive examination. Romeo labels this two-step logic 'procedural outsourcing,' and traces it across several member-state instruments, including prior bilateral arrangements Hungary struck with Serbia. The European Policy Centre has documented the trend in successive briefings on the externalisation of EU migration management.

The European Parliament and the Council of the European Union have acknowledged the legal fragility. The original Procedures Directive applied the 'safe third country' concept as an exception. Recurrent legal challenges, including those that reached the Court of Justice of the European Union prior to 2024, narrowed the conditions under which the designation could be invoked. Romeo reads those rulings as confirming that the term applies case-by-case rather than generically. The Italian protocol's blanket application of the designation to groups of nationalities, not individual applicants, contradicts what European Union case law has been slowly tightening toward for nearly two decades.

The empirical record

The numbers from the Albanian experiment are unusually granular for a policy that has been marketed as a deterrent rather than measured as a process. The 2024 group of 12 was initially processed under Italian jurisdiction at the centre in Shengjin. Italian courts ruled against the 'safe third country' designation for the named countries of origin. Detention at the Albanian facilities was followed by repatriation to Italy for full asylum proceedings. The second group, arriving in February 2026, was transferred back with similar judicial reasoning. The pattern points to a policy with a fixed failure rate at the front door.

Cost considerations are now in the foreground. The Italian government has paid the fixed annual sum agreed in the protocol to Albania; the operating cost of the centres runs into tens of millions of euros per year, on figures reported in Italian parliamentary hearings in 2025 and 2026. For the price of a functional accelerated procedure, Italy has so far obtained the repatriation of processed detainees to Italian territory rather than to the third country. The deterrent signal intended by the bilateral arrangement has not been measurable in the published record.

Counter-narrative

The Italian government's framing is procedural and rule-of-law-based. Asylum seekers who cannot make a credible claim, the argument runs, are using routes that bypass orderly access, and bilateral arrangements with cooperating third countries are a measured response in a constrained Mediterranean. The Rome Declaration of 2024, signed by Meloni and Rama on the bilateral arrangement, cites burden-sharing and the prevention of irregular crossings as core objectives. From that vantage, the courts' reversals are not a failure of the concept. They are the concept working as the rule of law intends: peer review of executive shortcuts. The European Union's own framework is structured around the same premise.

The opposite reading is sharper. Once courts in a single member state have repeatedly found that the expedited designation is legally defective, the doctrine, as deployed, adds processing weeks, return flights, and detention days without serving the procedural purpose for which it was invoked. Romeo's research is consistent with the second reading while declining to declare the underlying policy dead: the same procedural logic, refitted with the country-specific evidence the courts have demanded, could in principle function. Italy's June 2026 request to the Court of Justice of the European Union for a preliminary ruling will be the next measure of whether the legal shell can be reshaped or whether it has already collapsed.

Stakes

The longer pattern sits above the Italian file. The EU Pact on Migration and Asylum, formally adopted in 2024, embeds 'safe third country' and 'safe country of origin' concepts into a union-wide framework with mandatory procedural components. Romeo reads the Italian file as a stress test of that architecture before the Pact's full application in 2027. If the bilateral protocol becomes the template, the operative question is whether the European Union will accept accelerated procedures whose compliance with European asylum law is decided after the fact by domestic courts, or whether the Pact itself will tighten the designation standard. Member-state alignment around the second option is uneven. Hungary's programme, cited in Romeo's study, has produced its own litigation; the European Court of Human Rights has issued findings against other member states for return-related violations in cases unrelated to the bilateral protocol, with the Grand Chamber of the European Court of Human Rights having handed down more than 50 judgments finding violations of Article 3 of the European Convention on Human Rights in deportation cases in the decade before 2024. The procedural outsourcing route, in short, has not delivered durable legal cover even where it has been carried out consistently.

What is left is a concept waiting to be rewritten. The procedural alibi may yet be turned into a working legal instrument, but only if designation is anchored in country-specific evidence, return-chain agreements are made publicly auditable, and courts retain a fast route to overturn expedited designations. Italy's Albanian protocol, the only such policy that has been applied to a meaningful number of people and adjudicated within a defined jurisdiction, suggests the European Union now has the precedent it needed. The next twelve months of rulings will tell whether that precedent is adopted or litigated away.

This publication reviewed the procedures directive, the bilateral protocol text, and the academic literature on safe third country through publicly available channels. The framing here distinguishes between the doctrinal category as it appears in EU instruments and the operational record produced by Italian courts; the two are not the same thing.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://www.europarl.europa.eu/legislative-train/theme-towards-a-new-pact-on-migration-and-asylum/file-asylum-procedure-directive-recast
• https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex%3A32013L0032
• https://www.europarl.europa.eu/doceo/document/TA-9-2024-0028_EN.html
• https://curia.europa.eu/juris/document/document.jsf?docid=277110