Blog by Mark Symes of the Garden Court Immigration Law Team
International protection used to be one of the fastest moving areas in all of English law. But, in recent times, it has been litigation on asylum procedures which has developed constantly. But now DS [2026] UKUT 413 (IAC) gives pause for thought to every advocate and representative who argues a protection appeal.
As Professor Guy Goodwin-Gill wrote in his seminal test, The Refugee in International Law, determining an asylum claim represents:
‘an essay in hypothesis, an attempt to prophesy what might happen to the applicant in the future, if returned … a reasoned guess as to the future, taking account also of the element of relativity between the degree of persecution feared (whether death, torture, imprisonment, discrimination, or prejudice, for example), and the degree of likelihood of its eventuating’.
Key to that exercise (in Anglo-Saxon countries, at least) is the notion of the standard of proof. The change to that standard, found in Nationality and Borders Act (NBA) 2022 s32, has been something of a slow burner. Partly, no doubt, due to the de facto pause on status determination that followed the Rwanda experiment, which slowed the progress of decisions and thus appeals for a long time.
The Upper Tribunal (UT) had a crack at it, in JCK [2024] UKUT 100 (IAC), but the very recent landmark precedent of DS gives the provision greater scrutiny and comes up with conclusions of great significance. As the UT has sagely noted, NBA 2022 represents a paradigm shift for the determination of Refugee Convention appeals. The once relatively straightforward task of simply assessing credibility, risk, protection and internal relocation has been replaced by a staged assessment prescribed by statute, governed by different standards at different points in the process.
The two seminal decisions as refugee law developed following the creation of the universal right of appeal in 1993 were Kaja [1995] Imm AR 1 and Karanakaran [2000] EWCA Civ 11. Both noted the notorious difficulty many asylum-seekers face in “proving” the facts on which their asylum plea is founded. The decisions both sought to give the quietus to disputes that had previously riven the old Immigration Appeal Tribunal (an example being Raymond Maddison Esq’s dissent in the former) and even animated the Court of Appeal (CoA). A division of the CoA in Horvath [1999] EWCA Civ 3026 observed that the time might be ripe to authoritatively resolve the question of what standard of proof applied. But, whilst that division posed the question, history decreed that it was the Karanakaran court that answered it.
The judgments of two of the great refugee law judges, Sedley and Brooke LJJ, in Karanakaran, have long been seen as one of the most prescient of protection precedents and DS emphasises that they remain the guiding light by which decision makers should navigate. Those Lord Justices were alert to the fact that the previous governing precedent of Kaja, had (contrary to the general thinking that was prevalent) not held that facts were to be proved to the reasonable degree of likelihood threshold. This being a point to which the submission advanced by myriad advocates, ever since that “lower standard of proof” applied to credibility assessment, paid little heed.
What mattered, however, was that decision makers appreciated: “The need to reach a decision on whether an appellant had made his case to a reasonable degree of likelihood, arose only on the ultimate evaluation of the case, when all the evidence and the varying degrees of belief or disbelief were being assessed … [adopting a 2-stage process of assessing credibility then risk would mean that] … any uncertainties in the evidence would be excluded at the second stage, and that this could not be right.”
As Sedley LJ put it, not for the first time, displaying both intellectual insight and elegance (seldom matched before, or since): “So it is fallacious to think of probability (or certainty) as a uniform criterion of fact-finding in our courts: it is no more than the final touchstone, appropriate to the nature of the issue, for testing a body of evidence of often diverse cogency.”
One of the most significant contributions to the learning on credibility assessment that DS makes is its explanation of the relationship between subjective fear and credibility. It explains the rationale behind the statutory criteria found in s8 of the Asylum and Immigration (Treatment of Claimants, etc.) Act 2004, viz that conduct inconsistent with fearfulness may be relevant to assessing refugee status.
It clearly shows that the Karanakaran principles should continue to apply to the underlying question of credibility, however, because the only questions which the statute stipulates are to be assessed on the balance of probabilities standard are those of subjective fear and Convention reason. Thus, it is that under the NBA 2022 regime, it is only the ultimate assessment of both subjective fear and Convention fear that must be proved on balance of probabilities. The underlying alleged facts may be determined according to the cogency of the evidence in the decision maker’s estimation. All relevant evidence should be considered, except matters ruled out because no credence can be attached to it whatsoever.
Additionally, in the world of international protection appeals, one often encounters rather stale citations of precedent that might make the reader briefly imagine themselves to have time travelled to the early 2000s. In reality, there have been numerous, much more recent, important decisions giving guidance on assessing facts, directly in the international protection field, such as MAH (Egypt) [2023] EWCA Civ 216 and SB (Sri Lanka) [2019] EWCA Civ 160, and summarising that field’s edicts in the analogous field of trafficking; TVN [2021] EWHC 3019 (Admin) and MVN [2015] EWHC 1942 (Admin).
But now DS joins their number, as a convenient repository from which to draw useful, if historic, principles of credibility assessment. Particularly, the fact that corroboration is not required in absent circumstances, where it can reasonably be expected. Shortly before the Office of the United Nations High Commissioner for Refugees (UNHCR) at last redrafts their famous Handbook for Determining Refugee Status, the extant version gets some timely citations too. When some of us first appeared in refugee appeals, the UNHCR Handbook was all there was by way of authority. And DS reminds us that UNHCR were the originating source for the benefit of the doubt principle.
DS also corrects an aspect of the decision in JCK, which had provocatively held that the existence of Convention reason should be taken at its highest, and thus suggested that that issue be simply accepted rather than evaluated by any particular standard of proof. Clearly, this in fact fell to be assessed in the same way as subjective fear via an ultimate assessment on balance of probabilities, based on the diverse cogency of the historical facts. One feels for the JCK Tribunal here.
The general practice for some years in Home Office refusal letters has been to start by identifying whether or not a Convention reason is established, in order to determine whether the case is to be assessed via Refugee Convention or Humanitarian Protection standards. In so doing, the case is taken at its highest even though the facts have not been assessed.
Probably, this is a legacy of the Common European Asylum System when the Refugee Convention was given clear primacy, and, thus, it was necessary to determine the potential presence of a Convention reason at the outset. But whatever the reason, JCK was led into error in its suggestion that Convention reason should be assessed at its highest – in truth it should be assessed on the applicable standard of proof, viz balance of probabilities. Its approach was understandable, but wrong.
Of course, refugee status is just the starting point of a full international protection determination. Having assessed refugee status entitlement on the basis of these statutory convolutions, the same international protection claim needs to be evaluated again outside that framework.
Whilst not every Humanitarian Protection claim triggers the Refugee Convention, given the Convention reason requirement, it would seem that every Refugee Convention claim does have a potential Humanitarian Protection claim running alongside it. This is because it is difficult to imagine a case involving a well-founded fear of persecution that does not, at the same time, manifest a real risk of serious harm.
The same goes for the compatibility of any removal from the UK with threats to the right to life and to be free from inhuman and degrading treatment, found in European Convention on Human Rights (ECHR) Articles 2 and 3. These claims must be assessed to the “real chance” threshold, to which the unadulterated principles of credibility assessment, laid down from Karanakaran onwards, apply with full force.
To instruct Mark Symes, please contact immigrationclerks@gclaw.co.uk










