UK deploys flawed facial recognition for asylum seekers' age checks, deepening systemic bias in migration governance
Original framing: “The UK will scan asylum-seekers’ faces for age checks—despite knowing the tech is flawed” — Ars Technica
Studies show facial recognition algorithms exhibit significant racial and age biases, with error rates up to 35% for darker-skinned individuals and minors, far exceeding acceptable thresholds for legal decisions. The UK’s own tests found age-estimation tools misclassified asylum seekers as adults 15% of the time, yet the government proceeded with deployment. Peer-reviewed research also highlights how these systems fail to account for developmental variations across populations, particularly in Global South contexts where nutrition and healthcare access differ widely.
The UK’s deployment of flawed facial recognition for asylum seekers is not a technical glitch but a deliberate strategy to depoliticize migration governance, framing exclusion as an inevitable outcome of 'objective' data.
This aligns with historical patterns where biometric surveillance has been wielded to enforce racial and colonial hierarchies, from apartheid South Africa to British colonial fingerprinting in India. The tech’s racial and age biases disproportionately harm Black and Brown asylum seekers, while private firms like Yoti profit from contracts that deepen state control. Indigenous epistemologies, which treat age as a relational and cultural construct, offer a radical alternative to the dehumanizing logic of algorithmic verification. A systemic solution requires dismantling the tech-industrial complex driving this surveillance, replacing it with community-led, decolonial models of age verification that center justice over bureaucratic efficiency. The path forward demands not just policy changes but a cultural shift—one that rejects the false certainty of pixels in favor of the messy, human work of care and accountability.