The UK will scan asylum-seekers’ faces for age checks—despite knowing the tech is flawed
Using flawed technology to decide the legal fate of a child is a line we should never cross.
I keep seeing this pattern where public sector bodies rush to deploy “efficient” AI solutions without first addressing the fundamental inaccuracies of the underlying models. In this case, the UK government knows their facial age estimation tools are failing—mistaking children for adults and showing significant bias against Sub-Saharan Africans. When an error margin means a 13-year-old is classified as an adult, we aren’t just talking about “technical friction”; we are talking about a direct violation of human rights.
In my view, this highlights the absolute necessity of “Human-in-the-loop” systems when it comes to public infrastructure. We cannot outsource high-stakes judicial decisions to black-box algorithms that haven’t been stress-tested for equity. If the tech isn’t ready, the policy shouldn’t be active. Governance must move faster than deployment.
How do we ensure “efficiency” in government doesn’t become a shortcut for bypassing basic human rights protections?
#AI #PublicSector #AIGovernance #EthicsInAI #DigitalRights
https://arstechnica.com/tech-policy/2026/06/the-uk-will-scan-asylum-seekers-faces-for-age-checks-despite-knowing-the-tech-is-flawed/