The Supreme Court's decision in A Reference by the Attorney General for Northern Ireland [2026] UKSC 16 (AGNI), handed down on 2 June 2026, has had huge ramifications across the health and social care sector. The Court overturned the Cheshire West "acid test" for determining whether someone is being deprived of their liberty, declaring it wrong in principle and never adopted by the European Court of Human Rights. A new multifactorial assessment must now be applied. 

Following the publication of the Department of Health and Social Care’s guidance on “what constitutes a deprivation of liberty” (15 June 2026), several Court of Protection judgments are now emerging which demonstrate how judges are applying the new multifactorial test in real-life cases. Together, these are shaping what deprivation of liberty now looks like in a health and social care context. 

What did AGNI decide?

In AGNI, the Supreme Court identified a prisoner confined to a cell as the paradigm example of a deprivation of liberty (DoL). It held that a person must be sufficiently close to that paradigm to be regarded as deprived of their liberty. As a result, significantly fewer people are likely to meet the threshold for a DoL following the judgment.

The Court affirmed that whether a person is deprived of their liberty requires a multifactorial assessment of the person's specific situation, taking into account the type, duration, effects and manner of any restrictions, with no single factor being determinative. 

Crucially, contextual factors previously treated as irrelevant under Cheshire West should no longer be excluded. These include an individual's compliance or lack of objection to their care arrangements, the relative normality of the circumstances and setting, and the purpose of the restrictions. 

The Supreme Court also made clear that even if a person lacks capacity under the Mental Capacity Act 2005 (MCA) to consent to their care and living arrangements, they may nevertheless be capable of giving "valid consent" to their confinement if they have a basic understanding of their situation and can communicate their contentment with it, meaning there is no DoL. Our previous article explored the judgment in more detail.

How have the courts been applying AGNI?

Salford City Council v AR [2026] EWCOP 45 (T2)

Judge Burrows applied the AGNI multifactorial test to AR, a 25-year-old man with a moderate to severe learning disability and autistic spectrum disorder living in a supported living placement. Despite AR’s "substantial" restrictions, including 24-hour 1:1 supervision, 2:1 staffing in the community, locked doors, window restrictions, CCTV monitoring and highly structured daily routines, Judge Burrows found AR not to be deprived of his liberty. The finding turned principally on the protective purpose of the restrictions, the relative normality of the placement, AR's community participation and the absence of institutional or custodial characteristics.  

The case is practically useful for two reasons. First, the judge endorsed a detailed checklist of factors to consider in DoL assessments. These include: possibilities for P to leave; degree of supervision and control; availability of social contacts; purpose of confinement; use of sedative medication; use of restraint and coercion; the relative normality of the arrangements; effects and manner of implementation and the individual's degree of understanding of their circumstances; and duration of the restrictions. 

Secondly, he summarised five overarching principles: 

1. Non-secure environments matter. Living in a family home, foster placement, supported living or ordinary home is highly relevant and may point away from a DoL. 
2. Purpose matters. Protective and therapeutic restrictions are relevant and distinguishable from punitive confinement. 
3. Relative normality matters. The more the placement and care plan resembles ordinary life for a person in P's circumstances, the less likely it is to constitute a DoL. 
4. Protective motivation matters. Although not decisive, restrictions imposed for care and safety rather than punishment weigh against a finding of a DoL. 
5. Objection and consent matter profoundly. A person may lack decision-specific MCA capacity yet still possess sufficient factual understanding to express valid consent for their confinement, meaning they are not deprived of their liberty. The person must however be genuinely content with or accepting of their arrangements – mere acquiescence may not suffice.

Re Mustafa (a child) [2026] EWHC 2159 (Fam) 

This case concerned a 15-year-old boy with Prader Willi syndrome living at a specialist residential school. His placement involved locked external doors, a locked kitchen at night, constant 1:1 supervision and physical restraint as a last resort. However, Mustafa had repeatedly said he liked living at the school, had never sought to leave, and had made significant progress there. Even though he was sometimes restrained to prevent him hurting himself or others, he appeared to understand this and said he felt “fine” about it. The court agreed that Mustafa understood and was happy in his placement and therefore gave valid consent to his confinement, meaning there was no DoL. 

The case confirms that, post-AGNI, a child who is not Gillick competent may nonetheless have sufficient understanding to validly consent to their confinement. The judge also highlighted that, even if Mustafa withdrew his consent to the restraint, the school could still potentially restrain him under powers conferred by both common law and statute. 

Essex County Council v ST [2026] EWHC 2328 (Fam) 

Essex County Council v ST concerned a 15-year-old girl in local authority care, living in a residential placement. The restrictions included 2:1 supervision day and night, physical restraint when necessary, and locked doors and windows. ST regarded the supervision as oppressive and objected to it. Mrs Justice Theis found that ST was being deprived of her liberty, particularly because of the level of supervision and the way the restrictions were experienced by ST. 

Comparing this case with Mustafa’s situation, the key differences appear to be how each young person experienced the restrictions: Mustafa repeatedly said he liked living at the school, whereas ST regarded the supervision as oppressive and objected to it.  The extent of social contacts, educational provision and family contact were also relevant. ST's ability to engage with peers was restricted, she had lost significant academic years, and family contact was limited, whereas Mustafa enjoyed a full programme of activities and meaningful family contact. 

Derby City Council v SM [2026] EWCOP 46 (T2)

Most recently, Derby City Council v SM explored the overlap between the objective and subjective elements of deprivation of liberty. SM is a young adult with severe learning disability, catatonia, major depression and autism, living in a single-occupancy supported living bungalow. Her restrictions include 2:1 support for all daily and social activities. Although SM can be aggressive to herself, others and her environment, staff are able to manage her behaviour through distraction without the use of restraint. 

SM had repeatedly expressed dislike of her accommodation and said she wanted to live elsewhere, but her views about where she wanted to live were changeable and at times unrealistic. All parties agreed that SM was not validly consenting to her placement.

The judge concluded that SM was not objectively confined. She had never attempted to leave or refused to return to the bungalow; in practice, she welcomed the support she received; her living arrangements would have been similar at home; and staff did not use restraint. Because the objective element was not satisfied, there was no DoL. 

Are P’s views decisive as to whether there is a DoL?

The accredited legal representative argued that SM’s views were decisive and that her lack of consent could transform the arrangements into a DoL. HHJ Sue Williscroft disagreed however, holding that this was not the law following AGNI.

The decision confirms that the effect of the restrictions on P, and the way in which they are implemented, are relevant to the objective assessment of whether P is being confined, but not determinative. They must be considered alongside the other factors in the multifactorial assessment. If P is not objectively confined, a lack of consent to their living arrangements does not, by itself, amount to a DoL.

The significance of Care Act advocates

Notably, the local authority had agreed to provide SM with a Care Act advocate going forwards, which may have provided some reassurance that SM's voice and interests would continue to be heard without the need for the full apparatus of deprivation of liberty authorisation.

Key takeaways

The emerging case law confirms that DoL assessments must now be genuinely multifactorial and highly individualised. Despite some initial expectation that the courts might seek to avoid applying the AGNI framework in full, the decisions to date have applied it consistently and with rigour, with outcomes turning on the specific circumstances of the case rather than any single determinative factor. 

Some key takeaways for practitioners:

• The Salford checklist offers a valuable and practical tool for structuring DoL assessments in this new era.
• How P experiences the restrictions and their views on them are relevant to the objective assessment of whether P is being confined, but they are not determinative. The fact that P does not consent to or like the restrictions does not automatically make the situation a DoL – many other factors must be considered.
• The importance of Care Act advocates is coming to the fore. Where P is found not to be deprived of their liberty, there will not be regular reviews of their situation. However, an advocate can ensure that P is still supported and that their views, wishes and rights are properly represented.