Given that the AGNI case started with a devolution reference from Northern Ireland, it is of interest that – so far – there have been no reported cases from that jurisdiction applying the case. One reason for this that the Northern Ireland DoLS framework applies more broadly than the Anglo-Welsh one (so the ‘community DoL’ cases heard by the Court of Protection largely do not arise), and that challenges to authorisations under that framework are heard by the Review Tribunal, which does not publish its judgments. However, we now have a Northern Ireland High Court decision which considers AGNI, in the form of the decision of Humphreys J in A Health and Social Care Trust v WM & Ors [2026] NIFam 23. The case concerned a 13-year-old boy, LR, who was the subject of a full care order and who resided in a property owned and managed by the Trust.
LR had a diagnosis of autism spectrum disorder, global developmental delay and a moderate to severe learning disability. Humphreys J described him (at paragraph 5) as a “vulnerable and challenging boy [… who] has a history of violent and aggressive behaviour towards staff and other residents. In particular, he makes makeshift weapons and threatens to kill or stab other people. Between August 2024 and March 2025, LR physically assaulted staff 69 times and other young people on 13 occasions.” A move to his current placement had led to improvements in LR’s presentation, although “[c]lose supervision is still required, normally on a 1:1 basis but at times this is required to be 2:1. Restrictions around his access to the kitchen are required for health and safety reasons. Physical restraint was used on one occasion in April 2026.”
Before determining whether those restrictions gave rise to confinement and hence a potential deprivation of liberty, Humphreys J set out a detailed exegesis of the law as it had evolved before and after AGNI, taking into account decisions up to and including that of Theis J in Essex County Council v MH [2026] EWHC 2328 (Fam). He noted that:
39. The contrast between the decision in BM and that in ST is illustrative of the difficulties first instance judges have in the application of the multifactorial approach set out in AGNI. The absence of a bright line rule means that different judges can and will reach different decisions in very similar cases. Each case is, of course, fact specific, but it is important in this area, relating as it does to the care of some of society’s most vulnerable individuals that the law is clear and predictable. Many decisions around the well-being of those individuals affected are made by social workers and other professionals who are not judges or lawyers but wish to ensure that those concerned are provided for in the safest, most appropriate and lawful manner.
Humphreys J identified the two ‘checklists’ of factors identified in MH and AR.
In relation to children, Humphreys J observed that:
47. In Re A-F (Children) [2018] EWHC 138 (Fam) Sir James Munby set out his “rule of thumb” in relation to Cheshire West as it applied to children:
“Inevitably, one has to proceed on a case-by-case basis, having regard to the actual circumstances of the child and comparing them with the notional circumstances of the typical child … the best I can do, by way, I emphasise, of little more than ‘rule of thumb’, is to suggest that:
(i) A child aged 10, even if under pretty constant supervision, is unlikely to be “confined” for the purpose of Storck component (a);
(ii) A child aged 11, if under constant supervision, may, in contrast be so “confined”, though the court should be astute to avoid coming too readily to such a conclusion;
(iii) Once a child who is under constant supervision has reached the age of 12, the court will more readily come to that conclusion.”
48. The subject matter of AGNI concerned adults, defined for this purpose as those aged 16 or over. Although children under 16 therefore fell outside the direct subject matter of the reference, the Supreme Court’s exposition of the autonomous meaning of deprivation of liberty under article 5 is not age-specific. In my judgment the same principles must inform the assessment of whether the arrangements imposed upon a child amount to a deprivation of liberty, whilst recognising the distinct legal and factual circumstances of childhood. It remains to be seen whether the rule of thumb identified in Re A-F continues to have utility.
On the facts of the case before him, Humphreys J considered that:
49. The restrictions in place in relation to LR are supervision, access to the kitchen and physical restraint where necessary. He lives in a residential setting with his own bedroom, and the restrictions imposed on him are relatively modest and are for his own well-being. They are not punitive in nature. Physical restraint has only been used once this year. He attends school and a local youth club. There are no particular activities which he is prevented from doing. He appears generally content in his placement although there have been issues with the other resident.
50. Considering these factors cumulatively, I am satisfied that they amount to restrictions upon LR’s liberty but not a deprivation of it. The degree of supervision is substantial but responsive to need, rather than being constant. The environmental restriction relates to one part of the home. Physical restraint is an exceptional feature of his care. LR participates in education and community activities and he generally regards the placement as his home and is content to remain there. The purpose and practical effect of the arrangements are to enable LR to live safely in the community rather than to confine him. The arrangements remain far removed in degree and intensity from the paradigm of the prison cell identified in AGNI.
Although Humphreys J considered that it was “therefore unnecessary and undesirable for the court to determine the subjective element of consent,” he did note that he considered that, whilst the (English) Court of Appeal’s analysis in J v Bath and North East Somerset Council [2025] EWCA Civ 478, drew heavily on Cheshire West, it remained good law such that (applying its principles, given that it does not have direct effect in Northern Ireland), it would not have been possible for the Trust to have consented to any confinement to which LR was subject.
Humphreys J then took the opportunity to express (obiter) views about the impact of AGNI:
55. The Supreme Court in AGNI emphasised that vulnerable individuals who do not meet the threshold for deprivation of liberty still enjoy protection in the form of their article 2, 3 and 8 ECHR rights. One must question whether the protection afforded by the other relevant Convention rights is real or illusory. There is no equivalence between the regular supervision of the court and the steps required to enforce such rights. It is important to recognise that many children in the care system are the offspring of vulnerable adults, who suffer from mental health or substance abuse problems, or themselves lack capacity. They may simply not be able to make complaints or bring proceedings in relation to violation of their children’s rights.
56. Any child who is the subject of a care order will engage with the Looked After Child (‘LAC’) review process. This is an important statutory safeguard but is not equivalent to independent judicial oversight. It is therefore a real concern that vulnerable children may be the subject of unwarranted restrictions in their lives without regular, effective and independent review.
57. Equally, many vulnerable adults do not have a means of communicating concerns about their ongoing restrictions or treatment. There must be a danger of a “once and for all” approach being taken to consent which does not allow for changes of mind and circumstance. It is to be hoped that the forthcoming revised Code of Practice addresses some of these important issues.
58. There is no longer a bright line rule or an “acid test.” Whilst the outworkings of AGNI may result in many fewer people ultimately being made the subject of deprivation of liberty orders or safeguards, there will be a very significant body of work to be undertaken to assess individuals whose arrangements may fall into this category.
Having in consequence of his analysis above reached the view that the Trust’s application should be dismissed, Humphreys J then set out a point which is frequently misunderstood in relation to the cases that are being decided in relation to AGNI:
60. I emphasise that the factual conclusion reached in this case should not be treated as establishing that materially similar restrictions will necessarily fall outside article 5. The ratio of AGNI requires a case-specific assessment of the particular individual’s circumstances and the nature, duration, manner and effect of the measures in question. Such scrutiny is important to ensure that vulnerable adults and children do not unjustifiably lose the protection which the law otherwise affords them.
