Can AI determine a child’s age? The legal questions behind the Home Office's use of Facial Age Estimation

Why does age matter? 

Whether an asylum seeker is treated as a child or an adult can have significant consequences. Children are entitled to additional protections, including appropriate accommodation, education and safeguarding support. For this reason, age assessments play an important role in the protections and benefits offered to asylum seekers.  

Where an individual’s age is disputed and documentary evidence is unavailable, decision-makers must determine whether the person is under or over the age of 18. An incorrect assessment can have profound consequences, particularly where a child is wrongly assessed as an adult and loses access to protections designed to safeguard vulnerable young people.  

Age assessments also engage wider legal and human rights considerations. Public authorities must ensure that decisions affecting children are compatible with their safeguarding duties and broader legal obligations, including the principle that a child's best interests should be a primary consideration. Given the potentially serious consequences of error, decision-makers must exercise particular care where uncertainty exists. 

Why are children’s organisations concerned? 

More than 100 organisations working with refugee and migrant children have expressed concerns about the proposed use of FAE. These organisations, coordinated through the Refugee and Migrant Children's Consortium, include well-known charities and organisations such as Refugee Council, The Children's Society, UNICEF UK and British Association of Social Workers. They have warned that relying on AI to estimate age could increase the risk of children being wrongly treated as adults and placed in adult accommodation, detention facilities or other inappropriate settings. 

These concerns highlight a fundamental weakness in the proposed use of facial age estimation. Many young asylum seekers have experienced trauma, conflict, exploitation, displacement or difficult journeys before arriving to the United Kingdom. Such experiences may affect a young person’s appearance and development, raising serious doubts as to whether facial characteristics alone can provide a reliable indication of age. 

Age estimation is inherently complex because physical appearance varies considerably between individuals. Factors such as ethnicity, genetics, nutrition, health conditions, trauma and differing rates of physical development can all influence how old a person appears. This raises legitimate questions about whether facial characteristics alone can ever provide a sufficiently reliable basis for assessing age. 

The concern is not simply that AI may occasionally make mistakes. Rather, it is that the consequences of those mistakes can be severe and life changing. A child who is wrongly assessed as an adult may lose access to protections and could be placed in adult accommodation or detention centres which are inappropriate for their age and circumstances.  

Sam Baron, Chief Executive of the British Association of Social Workers, has emphasised that trained social workers bring professional judgement that AI cannot replicate. He notes that two young people of a similar age may look alike, but their life experiences, maturity and wider circumstances can differ significantly. Social workers are able to assess those wider factors as part of a holistic assessment. Many lone child asylum seekers arriving in the UK are also aged between 16 and 17. According to Home Office data, young asylum seekers are more than twice as likely to be recorded as children following assessments by social workers than those carried out by immigration officers at the border, with more than two-thirds ultimately assessed to be minors. 

Scientific Advisors to the Home Office have also expressed concerns that the government rushed the adoption of AI for political reasons and disbanded a committee to avoid criticism of the technology’s inadequacies. Media reports have also highlighted concerns that the technology exhibits potential racial bias, with critics arguing that it performs less accurately for certain ethnic groups, including some migrants from Sub-Saharan Africa. If accurate, such limitations could significantly increase the risk of children being incorrectly assessed as adults.  

What does the law say about age assessments? 

The legal concerns surrounding facial age estimation become clearer when considered against the existing framework for age assessments. 

The judiciary has recognised that age assessments should not be based solely on appearance. In R (B) v Merton London Borough Council, the High Court emphasised that age assessments must take a holistic approach, considering a range of factors rather than relying simply on physical appearance. This landmark judgment established what is now known as a "Merton-compliant" age assessment, which requires trained social workers to carry out a fair, thorough and evidence-based assessment of a young person's age by considering their history, background, behaviour and wider circumstances, rather than relying on appearance alone. 

As practitioners will know, so-called “Merton-compliant” age assessments remain central to the current system. The purpose of this approach is to ensure that disputed age assessments are evidence-based, fair and capable of taking account of each individual's unique circumstances, rather than relying on assumptions based on appearance alone. This reflects a longstanding recognition that age cannot be determined reliably from appearance alone.  

Can FAE work within the existing framework? 

The proposed use of Facial Age Estimation raises an important legal question: can AI be used in a way that is consistent with the existing framework for age assessments? 

The courts have long recognised that appearance alone is an unreliable indicator of age and that assessments must take account of a young person’s individual circumstances. Yet facial age estimation is, by its very nature, based primarily on analysing appearance, which raises important questions about how it can be reconciled with the holistic approach required by the courts. 

Alongside questions about whether facial age estimation is compatible with the existing legal framework, there are also important concerns about transparency and accountability. If an individual disputes an age assessment, it is important that they are able to understand and challenge the basis upon which that decision has been reached. Where AI forms part of the decision-making process, questions may arise about how those assessments are explained, reviewed and challenged, particularly if the underlying technology is not fully transparent. The involvement of AI introduces additional obstacles into a process that is already highly sensitive for the asylum seeker. 

For these reasons, it is difficult to see how a system that seeks to estimate age from facial characteristics can comfortably sit alongside a legal framework that has consistently recognised the limitations of appearance-based assessments and has emphasised a more holistic approach to assessing age.  

Can an age assessment be challenged? 

Where an individual believes they have been wrongly assessed as an adult, it may be possible to challenge the decision. The appropriate route will depend on who carried out the assessment and the circumstances of the case. Depending on the circumstances, individuals may also be able to seek a further or independent age assessment, or rely on additional evidence to challenge the original decision. 

Age assessments undertaken by local authorities may be challenged through judicial review where there are concerns that the assessment was unlawful, procedurally unfair or failed to follow the principles established by the courts. Evidence such as identity documents, medical records, witness statements and expert evidence may all be relevant, although each case will depend on its own facts. 

Given the significant consequences that an incorrect age assessment can have for accommodation, education, safeguarding and access to support, it is important that young people receive specialist legal advice as early as possible where they believe a decision is wrong. 

What could the future hold for AI in age assessments? 

The Home Office has presented facial age estimation as a tool that can assist decision-makers, but any potential efficiency gains should not come at the expense of fairness, particularly where decisions affect vulnerable children. Age assessments are decisions that can determine whether a young person receives the protections afforded to children under the law. The introduction of AI into this area therefore needs to be carefully considered.  

Whilst technology may have a role to play in supporting decision-making, significant questions remain as to whether facial age estimation can be reconciled with the legal principles that currently underpin age assessments. The technology may continue to develop, but its use must remain consistent with the safeguards established by the courts and the wider legal framework designed to protect children. 

If facial age estimation becomes embedded within immigration decision-making, wider questions may also arise about whether similar technologies could be adopted elsewhere across public services. This makes it all the more important that appropriate safeguards, transparency and accountability are established from the outset. 

How we may be able to help 

Age assessments can have life-changing consequences for children and young people seeking asylum in the UK.  

Our Public Law and Human Rights advises children, young people and their families on a wide range of public law matters, including age disputes, challenges to unlawful decisions by public authorities and judicial review proceedings. 

If you or someone you know has concerns about an age assessment or another public law decision affecting a child's rights, our specialists can provide clear, practical advice about the options available. Call us on 0808 149 9561 or request a callback at a time that suits you, and a member of our team will be happy to discuss your situation.