Secret AI tools now run key UK immigration checks

The Home Office has integrated artificial intelligence into numerous immigration tasks without extensive public review, according to a report released by the Open Rights Group. The research, titled Automated Discrimination: Unpacking the Digital Hostile Environment, indicates that the technology now plays a role in identifying sham marriages, assessing the ages of asylum seekers, and inspecting the digital contents of new arrivals' phones.

The system also supports border surveillance through live facial recognition and drones. Internally, the department uses an opaque tool called Identify and Prioritise Immigration Cases, or Ipic, to triage and evaluate applications. The report notes that AI is employed to summarize interview transcripts and the country-specific policy evidence that informs those hearings. These findings were compiled from freedom of information requests, limited departmental publications, and questions raised by members of Parliament.

The authors argue that these digital tools are not neutral and often replicate existing societal biases. They contend that the technology is being trialed in near-total secrecy, with officials frequently citing national security to block document releases and redact details. The report criticizes successive governments for treating complex human migration issues as technical problems that can be solved by software, warning that this approach reduces individuals to data points.

The study has received cross-party support from Labour MP Bell Ribeiro-Addy, SNP MP Pete Wishart, and Green MP Siân Berry. Ribeiro-Addy noted that the Home Office has been reluctant to publish documents explaining the mechanics of AI in asylum processing or how accountability is maintained when tools fail. She questioned why vulnerable people with little political power should bear the greatest risks of these systems. Sara Alsherif, a program manager at the Open Rights Group, described migration as a human experience rather than a technical fix, calling for urgent transparency and strict regulation to limit the technology's spread.

The Home Office stated that AI serves as one of several tools for border protection and does not replace human decision-makers. A spokesperson explained that trained caseworkers make asylum decisions, using AI only to assist with navigating interview records and policy information. The department asserted that the tools are monitored through quality-assurance processes and cannot be used in isolation. The group also called for a public register of government AI tools and stronger freedom of information laws to address the current lack of oversight.

The Open Rights Group has urged the government to regulate AI and ban uses it deems undemocratic, such as the use of ChatGPT in asylum applications. The report states that relying on AI to determine who stays and who is returned to danger is a deeply alarming move. The department has committed to publishing updated evaluations of its AI rollout to address concerns about the digital hostile environment.