Home Office Trials AI Age Estimation for Asylum Seekers
The Home Office is piloting the use of artificial intelligence (AI) to estimate the ages of asylum seekers, a move that has sparked debate and raised concerns about accuracy. An opinion piece published on The Conversation platform, authored by Professor Oli Buckley, a cyber security expert at Loughborough University, examines the technology and its potential limitations.
The Home Office faces ongoing challenges in determining the ages of individuals arriving in the UK seeking asylum, particularly when official documentation is lacking. Current methods often rely on physical examinations and interviews, a process that can be time-consuming and subjective. The new AI system aims to provide a faster and more objective assessment, utilising facial recognition and other biometric data to predict age.
Professor Buckley’s article highlights the significant risks associated with relying on AI for such crucial decisions. He points out that AI algorithms are trained on datasets, and if these datasets are biased – for example, if they predominantly feature images of one demographic group – the AI’s predictions can be inaccurate for individuals from other groups. This raises serious ethical concerns, particularly when applied to vulnerable populations like asylum seekers, where misidentification could have profound consequences on their legal status and access to support.
The accuracy of the AI technology is also questioned. While proponents claim it can achieve a reasonable degree of accuracy, Professor Buckley argues that the margin of error remains too high to justify its use in determining age for asylum claims. He emphasises the potential for false positives and false negatives, which could lead to unfair and potentially devastating outcomes for individuals.
The implementation of this AI system follows a recent legal challenge where the Home Office was criticised for its age assessment methods. The government is under pressure to streamline the asylum process, but Professor Buckley’s analysis suggests that technological solutions should not come at the expense of fairness and accuracy. Further scrutiny and robust safeguards are needed to ensure that the use of AI in age assessment does not exacerbate existing inequalities and injustices within the asylum system.
The Conversation article encourages a cautious approach to the adoption of AI in this context, advocating for thorough independent evaluation and ongoing monitoring to assess its impact and prevent unintended harm.
