"There are substantial and well-founded concerns about the bias of FAE," the coalition of rights groups wrote, bluntly capturing why 62 organisations have urged the UK government to abandon plans to use AI-powered facial age estimation (FAE) on people seeking asylum.
62 rights groups tell the Home Office to halt FAE rollout
The open letter, sent to the border security and asylum minister Alex Norris, was signed by 62 organisations including Amnesty International, Human Rights Watch, Liberty, the Electronic Frontier Foundation, Foxglove, and the Open Rights Group. The groups demanded the Home Office abandon its plans to roll out facial age estimation from 2027 and gave the department 21 days to answer detailed questions covering testing methods, training data, safeguards, appeal mechanisms, and how facial age estimates would influence asylum decisions.
Home Office plans and the claim the tool will "support" human decision-making
The Home Office has proposed using AI-powered facial age estimation to help immigration officers determine whether people claiming to be children are likely to be over or under 18. Ministers have insisted the technology will support, rather than replace, human decision-making. The Home Office has released only limited details about its testing programme; The Register asked the Home Office to comment.
Bias, "baked-in failures", and the 16-to-18 boundary
The coalition's letter argues the technology carries "baked-in failures and discrimination," singling out women and people of colour as groups likely to be disproportionately affected. The groups pointed to the Home Office's own guidance showing the technology's performance varies by ethnicity and skin tone, and noted that asylum-seeking children are predominantly people of colour — a demographic detail the signatories said makes the technology's reliability especially questionable for this purpose.
The organisations also highlighted a practical technical limitation: the Home Office acknowledges that FAE systems are imprecise at the "crucial 16-to-18-year-old boundary." Citing government figures, the letter says even the best-performing systems have an error margin of roughly 2.5 years in that range — precisely where accuracy matters most for age disputes in asylum cases.
The groups warned that age estimation may perform worse among asylum-seeking children because "trauma, violence, malnutrition, dehydration, sleep deprivation, and long journeys can leave children looking older than they are," a list of conditions the letter said could skew algorithmic estimates and increase the risk of misclassification.
Questions about data, consent, and missing impact assessments
The coalition demanded transparency about the images and datasets used to develop and test the FAE systems, saying it is unclear how consent could lawfully have been obtained if asylum-seeking children were included. The groups noted that the Home Office has yet to publish detailed results, methodologies, or impact assessments that would permit independent scrutiny of the technology's performance.
Crucially, the letter also noted that no Equality Impact Assessment or Data Protection Impact Assessment has been made public — documents campaigners say are necessary to evaluate legal and privacy risks before a high-stakes deployment.
What this means for asylum-seeking children, ministers, and technologists
- Asylum-seeking children: Campaigners argue the greater risk is that vulnerable children will be treated as adults if FAE overestimates age, a consequence the letter frames as a direct threat to child protection.
- The Home Office and ministers: Faced with a coordinated letter from 62 organisations, the department has 21 days to respond on testing methods, datasets, safeguards and appeal mechanisms — an immediate procedural pressure point ahead of the planned 2027 rollout.
- Technologists and privacy advocates: The groups have called for publication of detailed testing results, methodologies and the datasets used, and have raised legal questions about consent and the absence of public impact assessments; until those materials are released, independent validation is not possible.
The row exposes a simple but consequential tension: ministers say the tool will help officers; campaigners say the risks of bias, inaccuracy at the pivotal 16–18 boundary, opaque datasets, and absent impact assessments make that claim premature. Until then, the government's AI age guesser "remains a technology it says works, but has yet to fully show its workings."
