Section 72 “Danger to the Community” Requires Explicit Findings on Ongoing Extremist Support and Coherent Treatment of the Two Presumptions

1. Introduction

In A1 v Secretary of State for the Home Department [2026] EWCA Civ 807, the Court of Appeal (Singh LJ, with Lewis LJ and Edis LJ agreeing) allowed the Secretary of State’s appeal from the Upper Tribunal, holding that the First-tier Tribunal (“FTT”) had erred in law when it found that the respondent (“the refugee”) had rebutted the section 72 presumption that he constituted a danger to the community of the United Kingdom.

The appeal arose from the Secretary of State’s decision to revoke the refugee’s status under section 72(2) and (6) of the Nationality, Immigration and Asylum Act 2002 (“NIAA 2002”), following his conviction for terrorism-related offences and a sentence of two years’ imprisonment. The case was not a deportation appeal; the dispute concerned revocation of protection status under section 82(1)(c) NIAA 2002 and whether revocation breached the UK’s obligations under the Refugee Convention.

The key legal issue was how the FTT must reason when dealing with the two statutory presumptions in section 72(2) (particularly serious crime; danger to the community) and the rebuttal provision in section 72(6). The Court of Appeal’s central finding was that the FTT’s reasoning was legally defective because it did not confront a crucial factual question: whether the refugee still supported the extremist organisation whose propaganda he had intentionally disseminated, and it did not reconcile that question with the FTT’s own findings made under the first presumption.

2. Summary of the Judgment

  • The Court of Appeal held that the FTT committed a material error of law by failing to address a central factual issue relevant to the second presumption (danger to the community), despite making adverse findings under the first presumption (particularly serious crime).
  • The error was not treated as merely a reasons challenge; it went to failure to determine an essential fact where the burden lay on the refugee to rebut the presumption.
  • The Upper Tribunal erred in failing to identify that legal defect and in upholding the FTT’s decision.
  • The case was remitted to the FTT for redetermination because the missing fact-finding could not properly be completed by the Court of Appeal.

3. Analysis

3.1 Statutory and Convention framework

The legal mechanism was the Refugee Convention’s exception to non-refoulement in Article 33(2), reflected domestically by section 72 NIAA 2002 and the then-applicable Immigration Rules (including paragraph 339AC(ii)).

Under section 72(2), if a person is convicted in the UK and sentenced to at least two years’ imprisonment, there is a statutory presumption both that:

  1. the person has been convicted of a particularly serious crime; and
  2. the person constitutes a danger to the community of the UK.

Critically, section 72(6) makes the “danger” presumption rebuttable by the person. The burden therefore rests on the refugee to displace the statutory starting point.

3.2 The Court of Appeal’s core holding: inconsistency and the missing fact-finding

The Court of Appeal accepted that many criticisms of the FTT amounted to disagreement with evaluation rather than law. However, it identified a specific legal flaw: the FTT’s analysis of the two presumptions did not “join up”.

Under the first presumption, the FTT had made findings that the refugee:

  • had been convicted of intentional conduct (not merely reckless);
  • had been found in sentencing remarks to have a mindset supportive of the extremist group; and
  • had not acknowledged the intentionality of his acts or his support for that group, with aspects of his account inconsistent with the sentencing remarks and the probation material.

Yet when turning to the second presumption (danger), the FTT did not make a clear finding on the “crucial question”: whether the refugee still supported the extremist group at the date of the hearing, and if not, why not. The Court of Appeal held that this was a material error of law because it involved failure to determine an essential fact in an inquiry where the refugee bore the burden of rebuttal.

The Court also drew attention to the nature of terrorism dissemination: even if the offender does not intend attacks in the UK, online propaganda can inspire others. It treated this as reinforcing the need for explicit findings on whether the underlying mindset had changed.

3.3 Precedents cited and their influence

(a) EN Serbia v SSHD [2009] EWCA Civ 630; [2010] QB 633

The FTT itself cited EN Serbia v SSHD [2009] EWCA Civ 630 when considering “particularly serious crime”. The Court of Appeal treated EN (Serbia) as continuing to supply the key conceptual structure:

  • “Danger to the community” requires a real danger, and a real risk of repetition of a particularly serious crime will commonly satisfy that test.
  • As stated at paragraph 66 of EN (Serbia), once the State establishes conviction of a crime that is on its face particularly serious, the person will in practice need to show either: (i) the crime was not particularly serious due to mitigating factors associated with its commission, or (ii) because there is no danger of repetition he does not constitute a danger.

The Court of Appeal did not accept that EN (Serbia) collapses the two presumptions into one. Rather, it used EN (Serbia) to highlight the importance of a coherent risk analysis that confronts the likelihood of future serious harm, and the need for findings that genuinely answer the “danger” question.

(b) Brown v Rwanda [2009] EWHC 770 (Admin) and Brown v Government of Rwanda [2009] EWHC 770 (Admin)

The judgment adopted the established definition of “real risk” from Brown v Government of Rwanda [2009] EWHC 770 (Admin) at paragraph 34: a substantial risk, not merely fanciful, and not necessarily proved on the balance of probabilities. This framed the “danger” inquiry as one of practical assessment rather than mathematical probability.

(c) KD v Secretary of State for the Home Department [2026] EWCA Civ 349

The Secretary of State relied heavily on KD v Secretary of State for the Home Department [2026] EWCA Civ 349, but the Court of Appeal confined its relevance. It accepted as a general proposition (from paragraph 47 of KD) that dangerousness assessment involves: the nature of the crime, the likelihood of further serious offending, and the level of harm that might result.

The Court of Appeal also accepted the important implication for terrorism cases: given the potentially catastrophic harm, even a relatively low probability may still justify a finding that the statutory presumption is not displaced.

However, it rejected any reading of KD that would turn labels like “low risk” into an automatic legal conclusion, stressing instead that assessments are fact-sensitive and that KD was decided on perversity—a ground not advanced here.

(d) Ullah v Secretary of State for the Home Department [2024] EWCA Civ 201; [2024] 1 WLR 4055 and Volpi v Volpi [2022] EWCA Civ 464; [2022] 4 WLR 48

These authorities were used to underline appellate restraint: the Upper Tribunal can intervene only for error of law and should not “island hop” through evidence or demand exhaustive reasoning. The Court of Appeal accepted those constraints, but held that they do not excuse failure to decide a necessary factual question in a burden-of-proof setting.

The judgment also referenced AH (Sudan) v Secretary of State for the Home Department [2007] UKHL 49; [2008] AC 678 via Ullah, reaffirming deference to specialist fact-finding, while drawing the boundary where legal adequacy of the tribunal’s fact-finding is not met.

3.4 Legal reasoning: what the FTT had to do (but did not)

The Court of Appeal’s reasoning yields a practical rule for section 72 cases: where the tribunal’s own findings under the first presumption include adverse conclusions about intentionality, extremist mindset, or lack of acknowledgement, the tribunal must confront how (if at all) subsequent evidence demonstrates change sufficient to rebut danger.

In terrorism dissemination cases specifically, the tribunal must grapple with:

  • the possibility of indirect encouragement of attacks via publication (even if the offender did not “directly” intend UK attacks);
  • whether claimed protective factors (family life, employment, compliance with supervision, treatment engagement) actually bear upon ideological disengagement and propaganda-related risk;
  • the timeliness and scope of professional risk reports, and whether they address the precise risk features identified at sentencing.

Importantly, the Court did not rule out reliance on police monitoring or psychological/probation evidence; rather, it held that the tribunal’s reasoning must explicitly connect such evidence to the decisive question it must answer.

3.5 Impact

The judgment is likely to influence section 72 revocation (and closely related deportation) litigation in at least five ways:

  1. Coherence across presumptions: tribunals must ensure their treatment of the first presumption (seriousness) does not contain factual findings that are silently ignored when assessing danger.
  2. Mandatory engagement with “mindset” evidence: where sentencing remarks identify an extremist mindset, later decision-making must squarely address whether that mindset persists.
  3. Burden-of-proof discipline: because the refugee bears the burden to rebut the danger presumption, tribunals must not allow the inquiry to drift into an unstructured “low risk” conclusion without deciding the key factual propositions that would logically rebut danger.
  4. Terrorism dissemination is not confined to direct intent: the Court’s emphasis on indirect inspiration strengthens the analytical link between online propaganda offending and community danger.
  5. Remittal where the fact-finding gap is central: appellate courts are likely to remit rather than substitute decisions when the deficiency is failure to determine essential facts.

4. Complex Concepts Simplified

Section 72 presumption
A legal starting point that, if certain conditions are met (here, conviction + 2 years’ imprisonment), the law assumes both (i) a particularly serious crime and (ii) danger to the community. The person affected must disprove (rebut) the assumption where the statute allows.
“Danger to the community” and “real risk”
Not a requirement to prove future offending is more likely than not. A “real risk” is a substantial, non-fanciful possibility of serious harm.
Error of law vs disagreement on facts
Appellate bodies usually cannot re-run the factual evaluation. But it is an error of law if a tribunal fails to decide an essential issue, applies the wrong legal test, or reasons inconsistently in a way that undermines the legal conclusion.
Remittal
Sending the case back to the fact-finding tribunal for a new decision because the appellate court cannot properly fill the evidential or factual gaps itself.

5. Conclusion

[2026] EWCA Civ 807 clarifies that, in section 72 revocation appeals involving terrorism-related offending, a tribunal cannot lawfully find the “danger to the community” presumption rebutted without confronting and deciding whether the offender’s extremist support or mindset persists—especially where earlier findings (including the tribunal’s own findings under the first presumption) point to intentional conduct and extremist alignment. The decision reinforces disciplined reasoning under section 72, emphasising that “low risk” language must be anchored in explicit findings on the facts that actually rebut the statutory presumption.