Human Rights Watch (HRW) urged Britain’s Home Office on Friday scrapped its 2027 plans to use facial age estimation (FAE) AI to assess young asylum seekers, warning that the technology harbors discriminatory flaws and sets a dangerous global precedent for migration policy.
In one open letter To the UK Home Office, HRW and over 60 civil society groups raised significant concerns about discrimination in FAE, particularly in relation to women and people of colour. The FAE is an emerging technology currently being used in UK shops and bars to prevent minors from accessing age-restricted products such as cigarettes and alcohol. The human rights group said: “Using this to make life-changing decisions in refugee reception centers means introducing unreliable, untested technology into an already flawed process.”
The British Home Office emphasized that the use of FAE does not automate or replace what remains an integral human decision. Immigration officers remain responsible for taking several factors into account, such as a person’s behavior and appearance. Immigration officials use a “benefit of the doubt” approach to age disputes. They will treat a person as an adult only if two officers independently conclude that the person’s appearance and behavior clearly indicate that he or she is well over 18 years old. If they cannot meet this threshold, they treat the person as a child and refer them to a local authority for a more detailed, Merton-compliant age determination.
Nonetheless, Martha Dark, co-managing director of tech industry watchdog Foxglove, accused The government acknowledges the inaccuracy of its AI tools, saying internal government documents show the tools often misidentify children as adults. She further claimed that the systems were racially biased and exhibited particular accuracy problems when tested on people from countries in sub-Saharan Africa. Dark emphasized: “Children seeking asylum have often suffered unimaginable trauma. They should not be the test subjects for experimental technology that is embedded with inaccuracies and racial bias.”
A study by the National Institute of Standards and Technology (NIST) observed that the accuracy of age estimation algorithms varies depending on factors such as image quality and demographic characteristics, with performance varying by group. Higher error rates have also been consistently reported for female faces, but the underlying reasons are unknown.
