The UK Home Office says face age estimation will be an “additional” tool for border officers and that people will be treated as children if there’s uncertainty. Yet leaked testing from April 2025 found major deviations in the “best” algorithm, including risks
For people arriving after dangerous. physically demanding journeys across the English Channel. the first encounter at the border can decide what comes next. Now the UK is preparing to add face age estimation to that moment—even though internal testing flagged the technology’s performance gaps and bias.
The Home Office says face scanning is designed to be an “additional” tool for border officers and will not “replace or overrule human judgment.” A spokesperson also said that “In cases of uncertainty,” individuals “will always be treated as children until a further assessment is conducted.”
But when pressed on how the system will be used in real-world environments, the Home Office did not answer questions about its practical deployment.
The plan began with an announcement that the government would use face age estimation alongside border staff judgments to assess migrants in July 2025. Since then, the rollout has been delayed until 2027. Officials say the systems will use “cutting-edge AI tech” to “crack down on fake claims. ” with the aim of stopping “adults attempting to game the system.”.
That framing collides with what face age estimation has already shown elsewhere. Over the past five years. AI face scans have become part of controversial online age verification programs. as lawmakers have mandated social media platforms. porn websites. and some retailers check users’ ages. The technology has also been trialled at some bars and shops in the UK.
Face age estimation works by analyzing someone’s facial features, using systems trained on millions of age-labeled faces to produce an estimated age. In controlled laboratory tests, the best algorithms can predict a person’s age to within around 2.5 years.
In everyday conditions, the margins appear far less dependable. Results can vary wildly depending on the algorithm, a person’s gender, demographic details, and other factors. Poor-quality images—especially with bad lighting—can drastically reduce performance. The Home Office appears to have been aware of those weaknesses. including examples of people tricking some systems using images of characters from video games.
The leaked Home Office report. produced in April 2025 and completed before the government purchased face-scanning technology. describes how seven “FAE algorithms” were tested against more than 2.5 million images. The report says the unnamed “best performing algorithm” showed “substantial deviations” when tested on images of Sub-Saharan Africans. The report adds that. on average. the system also tended to predict that a 17-year-old would be over 18. and it performed worse on females.
Those findings land in a system that already relies heavily on human assessment. Tens of thousands of people make asylum claims in the UK each year. with many arriving after dangerous. physically demanding journeys in small boats crossing the English Channel. When border staff doubt that a person claiming to be under 18 is actually that age. the Home Office says officers can assess physical appearance. interview answers. and general demeanor to make an initial decision. These age estimations are made upon the “first encounter,” according to Home Office guidance.
Since 2010, official statistics say 40 percent of people who have faced age assessments have been classed as adults.
The leaked report also suggests that the testing environment may not reflect what will happen at the border. Its findings are based primarily on testing that uses high-quality images taken of documented people. which the report says may mean the algorithms’ accuracy rates would be even worse in practice.
The Home Office has indicated that FAE technology would help immigration officers making age assessments at the point of first encounter—precisely where the stakes are highest. where “uncertainty” is supposed to be resolved by treating someone as a child until further assessment. and where the technology’s known weaknesses could still shape the next step.
The government’s timeline keeps shifting. but the central tension stays the same: the system is being prepared to add a machine-derived judgment to a decision that. for some people. can determine whether they are treated as children at the start of their asylum process—even as internal testing points to errors that could move teenagers into adult territory.
UK Home Office face age estimation asylum seekers border officers AI age verification leaked report FAE algorithms 2025 2027 rollout cybersecurity and privacy concerns demographic bias immigration technology