The Upper Tribunal Has Started Answering the Questions the Home Office Hasn't

In my last piece on Facial Age Estimation, I set out the legal architecture the Home Office's AI tool must meet before it can be lawfully deployed at the border. The warning I ended on was this: the risk is that 2027 begins with a tool in operational use and a legal framework the courts will assess on the back of the first contested case.

That framing assumed the courts were waiting. They are not.

The Upper Tribunal has spent the last six months scrutinising the Home Office's age assessment architecture, not in the context of FAE, but in the context of the National Age Assessment Board. The NAAB cases are the clearest signal yet of where the Tribunal's analytical scrutiny is directed.

The National Age Assessment Board was established under the Nationality and Borders Act 2022. It allows the Home Office to conduct its own age assessments, independently of local authorities, in cases where the Home Office disputes the age claimed by an unaccompanied asylum seeker. Before 2022, age assessments were the exclusive province of local authority social workers, conducted under the Merton framework. The NAAB changes that. For the first time, the Home Office can assess age directly.

That matters for FAE. The logic of the system as designed is this: FAE informs the initial border decision; the NAAB is available for more comprehensive assessment where there is dispute. The question the UT has now been asked is whether the NAAB's assessments are legally sound. The answer, in the first successful challenge to reach judgment, is that they are not immune from being wrong, on the facts.

The facts of SS's case are worth setting out in some detail, because they illustrate, with unusual precision, the gap between what an age assessment can record and what it can miss.

SS was an Afghan young person who arrived in the UK in October 2022. Not knowing his own age or date of birth, he gave the Home Office a random date placing him at 17. He was initially accommodated in a hotel for unaccompanied minors. In January 2023, he was transferred to adult asylum accommodation when he was deemed to have turned 18. He later obtained an Afghan Tazkira recording his birth year as 2007, making him 16. He was referred to the London Borough of Croydon by the charity Young Roots, who took him into care.

Croydon referred SS to the NAAB under section 50 of the Nationality and Borders Act 2022. The NAAB assessed him between April and June 2024 and concluded he was an adult of 19, born in January 2005, two years older than he claimed. On the strength of that assessment, support and accommodation provided by Croydon was terminated. SS was rendered homeless: he had been granted refugee status before the assessment was completed and was therefore ineligible for asylum support. At the point the Upper Tribunal was hearing his case, he was being treated as an adult in the Youth Court.

A child was rendered homeless and processed through the adult criminal justice system as a direct consequence of an age assessment the Upper Tribunal later found was factually wrong. These consequences are not administrative. They are immediate and sometimes irreversible.

The NAAB's reasoning turned heavily on demeanour and credibility. SS had said 'I don't know' at least 158 times during the assessment, and 'I can't remember' no fewer than 49 times. The assessors concluded this reflected deliberate evasion, an attempt to conceal his true age. The Tazkira was dismissed as produced late and, in the assessors' view, created to support his claimed age. His initial reporting of a 2005 date of birth, which SS explained as a random date given in ignorance of his own age, was treated as an admission against interest.

What the NAAB assessors did not have was the medical and psychological evidence that became available for the Upper Tribunal hearing. That evidence established that SS had untreated epilepsy and PTSD, as well as symptoms of an adjustment disorder and significantly below-average cognitive functioning. Three medical experts confirmed the diagnoses. Professionals who had worked with him gave evidence that he presented as a child, with cognitive and emotional functioning more aligned with that of a young adolescent.

The Tribunal's finding was direct: SS's repeated inability to recall and his qualified, reluctant answers were explained by his untreated medical conditions, not by deliberate concealment. The assessors' conclusion of evasion was also, the Tribunal found, inconsistent with SS's actual conduct. He had provided his Tazkira. He had consented to his family being contacted. He had participated throughout a lengthy process. That is not the conduct of someone determined to obstruct.

The assessment was quashed. The Tribunal found, on the balance of probabilities, that SS was born on 1 January 2007, a child at the date of the NAAB's determination.

The Tribunal also rejected the Home Office's submission that NAAB assessments should in general be afforded greater weight than local authority assessments, given NAAB's independence and specialist training. That argument was firmly refused. The task of the Tribunal, to assess each case on its own merits, without any presumption as to age, is not altered by the NAAB's institutional status.

In April 2026, in R (NXP and NAN) v Secretary of State for the Home Department, the UT gave further procedural guidance, confirming that challenges to NAAB assessments belong in the Upper Tribunal, not the High Court, and setting out the framework for those claims.

There is a lesson in SS's case that extends well beyond NAAB. It applies to every age assessment conducted by any body under the Merton framework.

The assessors made a credibility finding, deliberate evasion, on the basis of behaviours that were fully explicable by medical conditions they were unaware of. That is not a failure of bad faith. It is a failure of information. The 158 'I don't know' responses were a diagnostic indicator, not a credibility indicator. Without the medical evidence, the distinction was invisible.

Before drawing any adverse inference from a young person's presentation, apparent evasiveness, inconsistency, poor recall, ask what else might explain it. Has a referral for psychological or medical assessment been made? Is there evidence of trauma, cognitive difficulty, or untreated health conditions? Is the young person communicating through an interpreter in circumstances where precision is structurally constrained?

This is not a counsel of perfection. It is a Merton requirement. The holistic assessment framework mandated since 2003 exists precisely because age cannot be determined from demeanour alone. SS is a concrete illustration of what happens when it is.

The NAAB cases are often treated as a matter for immigration practitioners. That framing is incomplete. Local authorities remain the primary holders of the statutory duty to safeguard children, and the interplay between NAAB assessments and local authority duties is the terrain on which these cases are fought.

When the NAAB assesses a young person as an adult, that person is diverted away from local authority children's services. In SS's case, the diversion was from Croydon's care to homelessness. Where a local authority has its own Merton-compliant assessment reaching a different conclusion, the legal tension is direct and live. The UT's willingness in R (SS) to examine the factual basis of the NAAB's determination, and to prefer a different conclusion, is a signal to local authorities that their own assessments, properly conducted and well-documented, carry weight in this jurisdiction.

That weight depends entirely on the quality of the assessment. R (SS) was decided on the facts. The Tribunal found the NAAB wrong on the evidence. That standard applies to local authority assessments too. The question is whether your recording demonstrates holistic, welfare-led reasoning, or a checklist. SS shows which one the Tribunal will scrutinise.

FAE is designed to inform the initial border decision. The NAAB is positioned downstream as the more comprehensive safeguard. But R (SS) demonstrates that the downstream safeguard is challengeable, and the UT will examine it on the merits. The chain from FAE to NAAB to UT challenge is shorter than the governance documents imply.

For a young person like SS, arriving without documentation, with trauma and untreated medical conditions that affect presentation, the risk at every stage of that chain is the same: that a behaviour which is a symptom is read as a sign of age or evasion. FAE will not detect epilepsy. A NAAB assessor, as R (SS) shows, may not either. The local authority Merton assessment, conducted holistically and with appropriate expert input, is the place where that risk can be identified and addressed.

If not, this is the moment to review it.