The United Kingdom is moving to adopt artificial intelligence (AI) facial recognition technology to identify adult immigrants posing as children. While some say the technology could prevent abuse of the system, critics warn that it "could violate children's rights."
"Determining age from a single photo?"…Why adopt it?
According to Yonhap News Agency and the BBC on the 29th, the UK Home Office recently signed a contract with an IT company to develop and test AI-based facial age estimation technology. The technology analyzes facial photographs taken during border screening to estimate an applicant's age.
Behind the British government's move to adopt such technology is the assessment that a significant number of adult immigrants claim to be minors. According to Home Office data, 43% of the 6,400 immigrants who identified themselves as minors at the border during the year ending March were confirmed to be adults.
Adult immigrants lie about their age because being recognized as children allows them to receive support under the protection system instead of going through asylum screening. Minors who enter the country without guardians are known to be provided with accommodation and welfare support, making their stay relatively easier.
Alex Norris, the UK's Minister for Border Security and Asylum, said, "Adult immigrants have abused the system by lying about their age, dispersing support that should go to children who actually need protection." He added, "We will use AI technology to quickly identify cases of system abuse and provide appropriate protection to children who need help."
Pilot operation from next year…Concerns over 'child to adult' misidentification
Currently, the UK determines age through professional immigration screening officers who review documents and conduct interviews, X-ray examinations, and MRI scans. However, as cases lacking documents or where identity is difficult to confirm increase, calls have been raised for faster means of determination.
The British government plans to use AI facial recognition technology as a tool to support screening officers' judgments. It plans to pilot the technology at the Dover refugee processing center in southeast England starting next year before reviewing whether to expand it. The Home Office explained that initial test results confirmed a certain level of accuracy and performance.
However, human rights groups are raising questions about the technology's reliability. In fact, last year's report by the UK government's immigration inspector confirmed not only cases of adults being misclassified as children, but also cases of minors being mistaken for adults.
The report pointed out, "There is no perfect method of age verification," adding, "Some misjudgments are inevitable, but they can become a serious problem, especially if they lead to depriving children of the protection and rights they should receive."
Anna Bacciarelli, senior AI researcher at Human Rights Watch (HRW), criticized, "Using inadequately verified technology in the process of determining whether a child is entitled to protection is dangerous," adding, "It is not even clear how accurate technology that estimates age from the face alone actually is."
