The AI age-detection tool being introduced by the British government threatens to misidentify child refugees as adults, according to a coalition of charities and human rights groups. The system, developed by German firm Cognitec, uses facial age-estimation technology to screen migrants at borders. Critics warn that racial bias in the algorithms will disproportionately affect Black children arriving from conflict zones such as Sudan and Somalia.
How the AI Age-Detection System Works
The Home Office announced a contract in May 2025 to deploy facial age-estimation technology. The system analyzes facial features to estimate a person's age, but it has a reported margin of error of up to 30 months. For minors, even a two-year overestimation can mean being treated as an adult.
Get the #1 Wireless Door Camera
REOLINK Bestseller: 2K Weatherproof Video Doorbell, No Monthly Fees.
Bias Against Sub-Saharan Africans
Testing of the Cognitec system revealed significant bias against sub-Saharan Africans, frequently overpredicting the ages of minors from that region. Maddie Harris of the Humans for Rights Network stated that children from these backgrounds are already “adultified” by the immigration system. The AI tool risks amplifying existing discrimination.
Consequences of Misidentification
Children wrongly classified as adults face grave dangers. They may be placed in detention centers, subject to removal proceedings, or housed alongside unrelated adults in asylum accommodation. Unaccompanied minors are especially vulnerable to violence and exploitation.
According to the Helen Bamber Foundation, 755 children in 2025 were incorrectly identified as adults upon arrival in the UK, based on Home Office figures. Harris emphasized that introducing AI is a “cheap process” that will not improve these numbers. “This is not about protecting children – this is about shoring up the decisions that [the authorities] are making,” she said.
Comparison of Age-Estimation Systems
| System | Error Margin | Bias Against Sub-Saharan Africans |
|---|---|---|
| Cognitec AI System | Up to 30 months | High (overpredicts age) |
| Manual Assessment (Immigration Officers) | Varies by training | Moderate (known racial bias) |
| X-Ray Bone Age Analysis | 6-12 months | Low (but invasive) |
Key Takeaways
- AI age-detection tools can overestimate age for children from sub-Saharan Africa, leading to wrongful adult classification.
- The UK government’s system has a 30-month margin of error, which is particularly risky for minors.
- Misidentified children face detention, removal, and placement in adult accommodations, exposing them to harm.
- Charities urge the Home Office to reverse plans and adopt more accurate, less biased methods.
FAQ
What is the AI age-detection tool used by the UK Home Office?
It is a facial age-estimation system developed by Cognitec that analyzes facial features to determine a person's age. It is being introduced to screen migrants at borders.
Why is the AI system biased against child refugees?
The system tends to overpredict the ages of minors from sub-Saharan Africa due to flawed training data and racial bias in the algorithm. This has been confirmed by testing.
What are the risks of being misidentified as an adult?
Children may be detained, removed from the UK, or housed with unrelated adults, exposing them to violence, exploitation, and trauma.
What can be done to protect child refugees?
Charities call for scrapping the AI tool, using more accurate age-assessment methods (such as medical evaluations), and ensuring that children are treated with appropriate safeguards during the asylum process.
As the UK government moves forward with this technology, the voices of human rights advocates grow louder. They demand that the system be halted until it can guarantee the safety and accurate assessment of every child refugee. Without change, more children will be thrust into the adult system—with potentially life-altering consequences.