To ask the Secretary of State for the Home Department, what assessment she has made of the accuracy of Facial Age Estimation technology procured by the Home Office when applied to children and young people seeking asylum.

To ask the Secretary of State for the Home Department, what alternatives were used to test against Facial Age Estimation technology in assessing the age of asylum-seeking children aged 15, 16 and 17.

To ask the Secretary of State for the Home Department, whether her Department has examined the relative margin of error in Facial Age Estimation technology as compared with other methods of age assessment, when it comes to the assessment of age of people seeking asylum.

To ask the Secretary of State for the Home Department, whether Facial Age Estimation technology procured by the Home Office has been tested on people seeking asylum.

To ask the Secretary of State for the Home Department, what assessment she has made of potential bias and differential error rates in Facial Age Estimation technology procured by the Home Office by ethnicity, skin tone, sex and country of origin.

To ask the Secretary of State for the Home Department, how decision-makers will account for uncertainty when the estimated age as per Facial Age Estimation technology falls near the threshold between childhood and adulthood.

To ask the Secretary of State for the Home Department, what datasets, images or other personal data have been used to train, validate or test the Facial Age Estimation technology intended for use in asylum age assessments.

To ask the Secretary of State for the Home Department, how Facial Age Estimation results will be presented to caseworkers or other decision-makers, and what weight those results will carry in any age assessment, including in relation to: accommodation, safeguarding, detention, removal and asylum-related decisions.

To ask the Secretary of State for the Home Department, what training and guidance will be provided to officials using or interpreting Facial Age Estimation results, including relating to the treatment of uncertainty and trauma-related facial ageing.

To ask the Secretary of State for the Home Department, what process will be available for a person seeking asylum to obtain, scrutinise, challenge or appeal a Facial Age Estimation result and any decision based on that result.

Assessing age is a complex task, and there is no single or a combination of techniques that are able to determine chronological age with precision.

Previously, the Department explored other potential scientific and technological methods, such as the use of X-ray and MRI technology. However, at this time FAE offers the potential to give much quicker results than other methods, with no requirement for any intrusive or invasive medical procedure. Our published factsheet, Facial age estimation - GOV.UK, explains how the technology works, and how the Department is exploring its use.

No final decisions have been made as to how FAE will work operationally, including how results will be displayed and weighted. The Home Office is planning a series of trials and testing throughout 2026 which will continue to inform policy, design and operational development. Subject to the results of the testing and trials taking place throughout 2026, FAE is planned to be rolled out from 2027. Full guidance will be provided and published to ensure a consistent approach to FAE and appropriate weighting of the results.

The Home Office has recently announced that the procurement process for an FAE algorithm has concluded, and a new contract has been awarded to Akhter Computers Ltd, with Cognitec as sub-contractor. The awarded algorithm was trained on the supplier’s proprietary dataset consisting of millions of facial images collected over several decades, covering a broad range of ages, ethnicities, and genders. The core model was not trained on Home Office operational data.

Separately, the Home Office has used industry algorithms to test FAE performance against approximately 2.5 million operational images with a known age across different ethnicities, genders and age ranges, and showed, indicating promising performance and accuracy. Internal testing and trialling is ongoing using the newly procured algorithm. Demographic differentials and the public sector equality duty have been a crucial part of this testing and ongoing policy development.