UK plans to use AI age checks on asylum seekers are drawing sharp criticism
The UK government’s decision to introduce AI facial age estimation for young asylum seekers whose ages are disputed has triggered a backlash from child refugee advocates, who argue the technology could push vulnerable minors into adult detention or prison settings.
The controversy follows a Home Office contract to roll out the system, which would be used in cases where officials question whether a young migrant is under 18. A coalition of more than 100 refugee children’s organizations says the move risks replacing contested human judgment with machine error in one of the most consequential decisions an asylum process can make.
Critics say trauma and migration history complicate automated estimates
According to a report from the Refugee and Migrant Children’s Consortium, age assessment is especially complex for unaccompanied young people seeking asylum. Many may have experienced trauma, under-nutrition and dangerous journeys before reaching the UK. Those factors, the group argues, can make appearance-based age estimation less reliable, particularly if systems are trained on norms that do not reflect the people being assessed.
The report does not reject the technology outright. Instead, it warns against relying on AI as a substitute for comprehensive assessments by trained social workers. That distinction is important. The core criticism is not simply that AI may be imperfect, but that its imperfections could be treated as authoritative in situations where the cost of a wrong decision is exceptionally high.
If a child is misclassified as an adult, the result could mean placement in adult accommodation, detention centers or other settings not designed to protect minors. For advocacy groups, that possibility makes safeguards essential rather than optional.
Data already show different outcomes depending on who performs the assessment
One of the most significant details in the debate is that young asylum seekers are reportedly more than twice as likely to be recorded as children when assessed by social workers than when assessed by immigration officers at the border. Home Office data cited in the source text indicate that more than two-thirds were assessed to be minors in social worker-led evaluations.
That gap suggests the process is not merely technical. It is shaped by methods, context and institutional incentives. Critics argue that adding AI to a border control system without strong limits may reinforce the harsher side of that divide rather than improve fairness.
The Home Office has emphasized concerns about adults making false claims to gain the protections afforded to children. That framing reflects a longstanding tension in asylum policy: preventing abuse of the system while still protecting people who may be especially vulnerable. The current argument is over whether AI can help balance those goals or whether it will make the tradeoff worse.
Advocates want AI treated as advisory, not decisive
The consortium’s report urges the government to use AI, if it is used at all, only in an advisory capacity. It also calls for safeguards including access to an appropriate adult, legal advice and a right to challenge decisions. Those demands reflect a broader principle in high-stakes public sector AI: automation should support human judgment, not close off routes of appeal.
That principle matters because age estimation is not a neutral administrative step. It shapes housing, legal treatment, welfare access and overall safety. In asylum systems, where documentation may be missing and trauma may affect how a person presents, advocates say any technology used must be handled with extreme caution.
The dispute also fits a larger pattern across governments experimenting with AI in migration control. Officials are drawn to tools that promise speed, consistency and efficiency. Civil society groups tend to ask whether those gains are real, what assumptions the systems encode and who carries the risk when the system gets it wrong.
For now, the UK case is less a story about a mature, settled technology than about a policy test with unusually high stakes. The Home Office is moving ahead with facial age estimation, while a broad coalition of refugee and children’s groups is warning that a tool built for disputed cases could end up deciding too much. The central question is whether the system will remain a narrowly used aid or become a shortcut that shifts children into adult systems before they can effectively contest the call.
This article is based on reporting by The Guardian. Read the original article.
Originally published on theguardian.com
