By Zurayda Mayet | Mayet & Associates

Introduction

On 12 May 2026, the Constitutional Court handed down judgment in Director-General, Department of Home Affairs and Others v Irankunda and Another [2026] ZACC 18. By a seven-to-two majority, the Court held that the Refugees Act 130 of 1998 (the Refugees Act) does not confer on a person whose asylum application has been finally rejected a clear right to submit a subsequent application based on changed circumstances. The Supreme Court of Appeal’s contrary conclusion in [2024] ZASCA 87 was set aside.

The judgment is a significant development in South African refugee law. It is also a self-consciously narrow one. The majority decided what the Refugees Act, on its existing text, permits. It expressly declined to decide whether the Constitution or South Africa’s international obligations may require a mechanism for subsequent applications. That latter question is, in our view, the next frontier of constitutional litigation in this area, and it is the question to which we return at the close of this article.

What follows is an analysis of the legal architecture of the Refugees Act as it now stands; of the Irankunda judgment in its proper statutory and case-law context; of the policy backdrop against which the judgment was delivered; and of the issues the Court left to Parliament and to future litigation.

The Statutory Architecture

To understand what the Constitutional Court did and did not decide in Irankunda, it is necessary first to understand the relevant provisions of the Refugees Act with some precision. Three elements of the statutory framework are central.

Section 2 — the non-refoulement principle. South Africa’s adherence to the non-refoulement principle is not merely a matter of international obligation; it is incorporated directly into domestic law by section 2 of the Refugees Act. The section prohibits the refusal of entry, expulsion, extradition or return of any person to a country where that person may be subjected to persecution on listed grounds, or where his or her life, physical safety or freedom would be threatened on account of external aggression, occupation, foreign domination or events seriously disturbing or disrupting public order in either a part or the whole of that country. Section 2 is the constitutional and international-law anchor of the domestic refugee regime.

Section 21(1) — the application mechanism. Section 21(1) provides that an application for asylum must be made in person, in accordance with the prescribed procedures, to a Refugee Reception Officer at any Refugee Reception Office. The section establishes the gateway to the refugee determination process; everything else (the interview, the section 24 determination by a Refugee Status Determination Officer, the section 24A treatment of manifestly unfounded applications, the appeal process before the Refugee Appeals Authority, and the confirmation process before the Standing Committee for Refugee Affairs) proceeds from a valid application under section 21(1).

Section 1 — the “abusive application” definition. This is the textual provision on which much of the analytical argument in Irankunda turned. The Refugees Act defines an “abusive application” to include an application made “after the refusal of one or more prior applications without any substantial change having occurred in the applicant’s personal circumstances or in the situation in his or her country of origin”. The very existence of this category — premised, on its face, on the assumption that prior-rejected applicants might lawfully submit further applications, formed the dissent’s most powerful textual argument that the legislature must have contemplated some repeat applications.

These three provisions, read together, set up the question the Constitutional Court was called upon to decide.

The Facts and the Litigation

Ms Amina Irankunda and Ms Arava Niyonkuru are Burundian nationals who fled to South Africa between 2008 and 2012. Each applied for asylum under section 21(1) of the Refugees Act. The Refugee Status Determination Officer rejected both applications as manifestly unfounded. The rejections were confirmed by the Standing Committee for Refugee Affairs in February 2014 and December 2014 respectively. Neither rejection was challenged on review at the time, and the determinations accordingly became final.

In 2015, Burundi’s political position deteriorated sharply following then-President Pierre Nkurunziza’s announcement that he would pursue a constitutionally contested third term in office. The announcement triggered widespread civil unrest, violence, and a humanitarian situation that attracted significant international attention.

In 2018, on the basis of these changed conditions, both women sought to submit fresh applications for asylum, contending that they had become sur place refugees, that is, persons who did not originally qualify for refugee protection at the time they left their country of origin, but whose entitlement to protection had crystallised as a result of subsequent events. The Department of Home Affairs refused to accept the further applications, on the basis that the Refugees Act does not permit a second application after a final rejection of the first.

The matter proceeded through the High Court (Western Cape Division), which agreed with the Department’s position, and on appeal to the Supreme Court of Appeal, which did not. The SCA in [2024] ZASCA 87 held that asylum seekers whose initial applications had been rejected could submit subsequent applications, and that the failure to consider those subsequent applications constituted a reviewable error. The Department appealed to the Constitutional Court.

The Majority Judgment

Writing for a seven-judge majority comprising Mlambo DCJ, Kollapen J, Mathopo J, Mhlantla J, Musi AJ, Savage J and Tshiqi J, Kollapen J framed the central issue as “whether there is a clear right in the Act to submit a subsequent asylum application following the refusal of the first application”. The Court held that there is not.

Three strands of the majority’s reasoning warrant attention.

First, the conceptual distinction between first-time sur place claims and subsequent applications. The majority’s central analytical move was to draw a sharp distinction between (i) a person who has never previously applied for asylum, and whose claim, when it is first made, happens to be based on events that occurred after his or her arrival in South Africa, and (ii) a person whose earlier asylum application has been finally rejected, and who seeks to lodge a further application based on subsequent changes in country conditions. The first is a first-time sur place applicant. The second is a subsequent applicant whose claim happens to invoke sur place circumstances. The Refugees Act, on the majority’s reading, accommodates the former through the ordinary section 21(1) process. It does not accommodate the latter.

The textual significance of this distinction is that it preserves the sur place concept (which is integral to South Africa’s compliance with the 1951 Refugee Convention) while denying that the concept opens a back-door route to repeat applications. Sur place describes the basis for a claim, not a mode of application.

Second, the finality of the determination process. The majority emphasised that the Refugees Act establishes a single, comprehensive application and determination process, complete with internal appeal mechanisms (to the Refugee Appeals Authority) and external review mechanisms (under the Promotion of Administrative Justice Act 3 of 2000 (PAJA) and the common law). The process is structured to reach a final outcome. The Act does not, on the majority’s reading, contemplate a parallel mechanism for the relitigation of a final outcome on the basis of supervening events.

Third, the institutional argument. The majority was alive to the practical consequences of the contrary view. An open right to submit subsequent applications based on changed country conditions, in the absence of legislatively prescribed limits, would risk undermining the finality of asylum determinations as a structural feature of the refugee system. It would also create an asymmetry between asylum seekers in the South African system and applicants in comparable jurisdictions, most of which subject subsequent applications to threshold tests, procedural filters, or time bars.

Critically, however, the majority did not say that subsequent applications could never be permitted. Its conclusion was the narrower one that, if a mechanism for subsequent applications is to exist, it must be expressly created by Parliament through legislative amendment, and not by judicial interpretation of a statute that, on its plain text, does not provide one.

The Dissent

Nicholls AJ and Rogers J, writing jointly in dissent, reached the opposite conclusion. Three lines of argument animate their reasoning.

The non-refoulement principle as an interpretive tool. The dissent placed substantial weight on section 2 of the Refugees Act and the corresponding international-law principle. Where domestic law is capable of more than one reading, the dissent reasoned, the reading consistent with the non-refoulement principle should be preferred. A reading that exposes a person to forced return to a country in which he or she now faces a real risk of persecution, even though the person did not face that risk at the time of the original application, is in tension with the non-refoulement principle. The contrary reading, which permits a subsequent application in circumstances of materially changed conditions, gives full effect to the principle.

The “abusive application” textual hook. Of the dissent’s textual arguments, the strongest is grounded in the Refugees Act’s own definition of an “abusive application” being one made after the refusal of one or more prior applications, “without any substantial change” in personal circumstances or country conditions. The definition presupposes, on its face, that some applications made after the refusal of prior applications are not abusive, specifically, those made after a substantial change in personal circumstances or country conditions. If the legislature had intended that no subsequent applications were ever permissible, the dissent argued, the carve-out for non-abusive subsequent applications would be unnecessary. The structure of the definition is consistent only with a legislative scheme that contemplates subsequent applications, while disqualifying those made without substantial change.

The equality argument. The dissent identified what it characterised as an irrational outcome of the majority’s interpretation: a person who has never previously applied for asylum in South Africa, but who now seeks protection on the basis of changed conditions in his or her country of origin, may make a first application. A person who applied years ago, was rejected for reasons that may have been entirely sound at the time, but who now faces the identical risk on the identical changed-condition basis, is permanently barred. The two persons are similarly situated in the only respect that matters under refugee law, the risk they face on return and the difference in treatment is, on the dissent’s analysis, difficult to justify against the constitutional standards of equality (section 9) and dignity (section 10).

The Judgment in the Broader Case Law

Irankunda sits within a developing line of Constitutional Court and SCA jurisprudence on the right to access the asylum system. Two cases provide useful context.

In Ruta v Minister of Home Affairs [2018] ZACC 52, the Constitutional Court held that an asylum seeker who expresses an intention to apply for asylum must be permitted to apply, even where there has been delay in approaching a Refugee Reception Office, and even in the absence of an asylum transit visa. Ruta concerned the right to make an application; it did not address the right to make a second application.

In Scalabrini Centre, Cape Town v Minister of Home Affairs [2017] ZASCA 126, the Supreme Court of Appeal considered the closure of the Cape Town Refugee Reception Office and the constitutional implications of restricting physical access to the application process.

Read alongside these cases, Irankunda completes a particular conceptual picture. The right to make a first application is robustly protected (Ruta). The right of physical access to the system is constitutionally significant (Scalabrini). The right to make a second application after final rejection, however, is not located within the existing statutory framework, and (per Irankunda) cannot be located there by judicial interpretation. It must be created, if at all, by Parliament.

The Policy Backdrop: The Final Revised White Paper

Irankunda was decided weeks after Cabinet approved, in early April 2026, the Final Revised White Paper on Citizenship, Immigration and Refugee Protection. The Revised White Paper proposes the consolidation of the Citizenship Act 88 of 1995, the Immigration Act 13 of 2002 and the Refugees Act into a single piece of legislation, and proposes the introduction of a “First Safe Country” principle under which an asylum seeker would generally be expected to seek protection in the first safe country he or she enters.

The alignment between the judicial and policy positions is unlikely to be coincidental. The Department of Home Affairs has been clear, both in its public statements and in the Revised White Paper, about its intention to limit what it characterises as abuse of the asylum system. Irankunda reinforces an element of that policy posture — the principle of finality in asylum determinations, without itself engaging the broader policy questions raised by the Revised White Paper, including the constitutional questions raised by the First Safe Country principle.

For practitioners, two observations follow.

First, the legislative reform programme contemplated by the Revised White Paper is the obvious vehicle for the legislative mechanism the majority in Irankunda contemplated. If Parliament wishes to create a subsequent-application mechanism, it has the opportunity to do so as part of the consolidated legislation contemplated by the Revised White Paper. That legislation is expected to be introduced in due course; whether the Bill will, in fact, include such a mechanism remains to be seen.

Second, the constitutional questions raised by the Revised White Paper itself, particularly the First Safe Country principle and its compatibility with section 2 of the Refugees Act and the international non-refoulement obligation, are matters on which further litigation is likely. Irankunda is therefore a relatively early instalment in what is likely to be a sustained period of constitutional engagement with the architecture of South African refugee law.

What the Court Did Not Decide

It is worth being precise about the scope of the judgment, because the boundaries the Court drew matter for what comes next.

The majority decided that the Refugees Act, on its existing text, does not confer a clear right to submit a subsequent application after final rejection. It did not decide:

• whether section 9 (equality) of the Constitution may require the creation of such a mechanism;
• whether section 33 (just administrative action) bears on the procedural treatment of a person whose claim has, in substance, changed since the original rejection;
• whether section 2 of the Refugees Act, read with South Africa’s obligations under the 1951 Refugee Convention and the 1969 OAU Convention, may require the creation of such a mechanism as a matter of statutory interpretation in pari materia with constitutional rights and international obligations;
• whether the prohibition on subsequent applications, as a matter of practical application, may yield to the non-refoulement principle in particular cases where the risk on return is concrete and verifiable;
• whether the current section 24A procedure for “manifestly unfounded” applications, when properly applied to a first-time sur place claim made by a person who has not previously applied, can adequately discharge South Africa’s protection obligations in the Irankunda fact-pattern.

Each of these questions is, in our view, available for litigation in an appropriate case.

Implications for Practice

Three practical observations bear on the conduct of asylum and refugee matters in the wake of Irankunda.

First, the importance of comprehensive initial applications. The judgment underlines the practical significance of properly preparing a first-time asylum application. Where a claim has both a pre-departure component (events in the country of origin before the applicant left) and a sur place component (events occurring after the applicant arrived in South Africa), both should be advanced in the original application. Irankunda increases the cost of a deficient first application materially, because the route to a corrective second application has, on the majority’s reading, been closed.

Second, the importance of properly pursuing internal and judicial review where the first decision is wrong. Where a Refugee Status Determination Officer rejects an application, the route to a corrected outcome lies through the Refugee Appeals Authority, the Standing Committee for Refugee Affairs, and (on review) the High Court under PAJA. Irankunda does not foreclose any of those routes; it confines applicants to them. Practitioners should be alive to the strict procedural timelines that govern internal appeal and judicial review, and to the importance of not allowing a wrong rejection to become a final rejection.

Third, the prospect of constitutional litigation on the questions the Court left open. Where a client’s circumstances raise, on the facts, one of the open questions identified above, particularly where the non-refoulement risk on return is concrete and supportable, the route to relief is no longer through a “subsequent application” treated as if it were a first application. It is more likely to be through constitutional litigation seeking to read in, or otherwise to obtain, a mechanism that the Refugees Act, on the majority’s reading, does not provide. The architecture of such a case will be different from a routine asylum review, and will turn on the strength of the factual record on country conditions and on the applicant’s personal risk profile.

Concluding Observations

Irankunda is a judgment that does what it does carefully. The majority resisted the temptation to decide more than the statutory question before it. The dissent set out, with equal care, the constitutional and international-law concerns that, in its view, ought to have led to a different conclusion. The result is a judgment that is, on its face, narrow, but whose practical effect is significant: the South African refugee framework now operates on the basis that there is one application, and one chance.

The judgment is unlikely to be the last word. The Court has, in effect, invited Parliament either to confirm that position through the consolidated legislation contemplated by the Revised White Paper, or to qualify it by creating a structured subsequent-application mechanism. It has also, in effect, invited constitutional litigation on the questions that the present judgment did not resolve. For now, the position is settled. For the medium term, the architecture of South African refugee law remains a moving subject.

How We Can Assist

Mayet & Associates advises on refugee and immigration matters, including the conduct of asylum applications and appeals, judicial review of refugee determinations, and constitutional litigation on questions arising under the Refugees Act, the Immigration Act and related instruments. Should your matter raise any of the issues discussed in this article, please contact the firm.

Mayet & Associates · www.mayet.law

This article is provided for general information only and does not constitute legal advice. Readers should obtain specific advice in respect of their particular circumstances before acting on any of the matters discussed.