top of page

Assurances have limits: Haxhia and the “right to hope” in extraditions to Albania

  • Writer: Giulia Borgna
    Giulia Borgna
  • Apr 30
  • 4 min read

In Haxhia and others v Republic of Albania [2026] EWHC 956 (Admin), 26 April 2026, the UK High Court gave an important ruling on the role (and limits) of assurances in extradition proceedings, as well as on the applicable standard in potential life-sentence cases in the requesting State.


The case concerned extradition requests issued by Albania in connection with an alleged murder. The requests had originally been certified on the basis of Article 78/a of the Albanian Criminal Code, namely murder due to blood feud. Under Albanian law, a conviction under Article 78/a may lead to life imprisonment without the possibility of applying for parole. The District Judge therefore accepted that extradition on that basis would create a real risk of treatment contrary to Article 3 ECHR.


a.      Assurances in extradition proceedings: role and limits


To cure that risk, Albania had provided assurances that the requested persons would not be prosecuted under Article 78/a, but under Article 78, i.e. premeditated murder. Unlike Article 78/a, Article 78 allows for the possibility of parole, although only after 35 years in the case of a life sentence.


The High Court held that this was not a legitimate use of assurances. The problem was not whether assurances are generally acceptable in extradition cases. They typically are, especially where they address prison conditions or other human-rights risks. The problem was that, in this case, the assurance did not merely reduce the risk attached to the original request. It effectively replaced the original extradition request with a different one.


That distinction is crucial.


As the High Court put it, “assurances are offered and accepted in order to facilitate compliance with extradition requests not as vehicles to change one request into a different request” (§ 26). Extradition is offence-specific: the requested person is surrendered for the offence identified in the request, not for a broadly similar factual narrative. In the circumstances of the present case, even if Article 78 and Article 78/a both concern murder, they are different offences under Albanian law. Consequently, the shift from one to the other was a matter of substance, not just form.


b.      The life sentence issue


The judgment is also interesting for another reason: the Court rejected the argument that a life sentence with review only after 35 years is automatically incompatible with Vinter v United Kingdom.


According to the High Court, Vinter does not impose a rigid 25-year rule. It requires that a life sentence be reducible in law and in practice: the prisoner must have a real possibility of review, allowing the authorities to assess rehabilitation and whether continued detention remains justified. But the reference in Vinter to review “no later than 25 years” was not treated as a “hard-edged standard (§ 62) applicable in every case.


This is an important point. The High Court accepted that Article 78/a raised a serious Article 3 issue because it excluded parole altogether. By contrast, Article 78 did not create the same problem, even though review would only become available after 35 years (the Hague Court of Appeal in 2019 reached a different conclusion, see here). For the High Court, the decisive distinction was between no realistic prospect of release and a late but legally available review mechanism.


The appeal ultimately succeeded, not because the 35-year review period was found to breach Article 3 ECHR, but because the certified extradition requests had been transformed during the proceedings.


c.       A short comment


The finding on assurances is persuasive: assurances may cure a risk, but they cannot rewrite the case. The finding on the 35-year minimum waiting period for the first review of a life sentence is, instead, more open to debate, while bearing in mind the specific nature of the assessment in the extradition context (see Sanchez-Sanchez v the United Kingdom [GC], 3 November 2022 and, more recently, Hayes and Others v. the United Kingdom, 1 July 2025; for a comment to Hayes, see here).


In T.P. and A.T. v Hungary, 4 October 2016, § 45, the ECtHR held that the 25-year period identified in Vinter and Others is not merely a general indication, but the “maximum recommended time frame after which the review of a life sentence should be guaranteed, established on the basis of a consensus in comparative and international law” (§ 26). Similarly, in F.B. and Others v the Netherlands, decided on 21 April 2026, the Court found no violation of Article 3 in the Dutch life-sentence system, where reintegration possibilities arose after 25 years from police custody or pre-trial detention, followed by an ex officio pardon decision after 28 years.


On the other hand, the Hungarian cases show the outer limit.


In Bancsók and László Magyar v Hungary (no. 2), 28 October 2021, the Court examined life sentences where parole could be considered only after 40 years, and found a violation of Article 3. The same concern appears in the very recent Éberling and Others v Hungary, decided on 14 April 2026, where the mandatory pardon procedure was available only after 40 years of a whole-life sentence. The Court again found that the system did not provide a Convention-compliant prospect of release.


The High Court in Haxhia rather summarily discounted subsequent Strasbourg case-law on the basis that those decisions were not delivered by the Grand Chamber, unlike Vinter (§ 62).


It is true that recent ECtHR case-law confirms that the decisive question is not only when review becomes available, but whether the sentence is reducible in law and in practice, with a genuine assessment of rehabilitation and continued penological justification. However, the 35-year threshold remains close to the outer edge of the Convention case-law. Strasbourg has accepted systems involving review around 25-30 years, while it has repeatedly rejected Hungarian mechanisms operating only after 40 years.


In conclusion, while Haxhia offers a clear answer on the impermissible use of assurances to reshape an extradition request, the compatibility of a 35-year waiting period with the “right to hope” remains very much open to debate. Further guidance from the European Court of Human Rights would be particularly valuable with regard to the Albanian legal system. To date, the only case to have reached the final stage was recently declared inadmissible because the applicant, although formally serving a life sentence, was a fugitive (Gjinarari v. Albania (dec.), 17 March 2026, §§ 13-14).



 
 
 

Comments


bottom of page