Artificial Intelligence

UK Home Office Faces Potential Probe Over OpenAI Tool Use

A UK government watchdog is weighing an investigation into the Home Office's deployment of AI systems to handle asylum applications, amid growing scrutiny over tool accuracy and government secrecy.

Radwan AbdulhadiEditor and technology writer

Published: October 28, 2026 · Updated: October 28, 2026 · 3 min read

UK Home Office Faces Potential Probe Over OpenAI Tool Use — Image: WIRED

Potential Watchdog Investigation

A UK government watchdog is considering an investigation into the Home Office’s deployment of generative AI tools to process asylum applications. The potential probe follows public questioning from lawmakers and researchers regarding the accuracy of the technology and the government's secrecy surrounding its use.

The Independent Chief Inspector of Borders and Immigration (ICIBI), which monitors the Home Office’s management of borders, immigration, and customs, has asked the digital rights nonprofit Open Rights Group (ORG) to share more information on the tools as it formulates its inspection priorities for next year. ORG’s findings stem from a six-month investigation involving 11 parliamentary questions, two ministerial letters, and 10 public records requests aimed at uncovering details withheld by the Home Office.

Pilot Program and Case Processing

The immigration and law enforcement authority began piloting the Asylum Case Summarisation (ACS) and Asylum Policy Search (APS) tools in late 2024. These tools were built on OpenAI’s 2023 GPT-4 large language model and Microsoft Foundry to summarize interview transcripts with asylum applicants and country-of-origin information.

The initiative was introduced as the Labour government attempted to clear a large backlog of asylum cases while managing tens of thousands of new applications received annually. Government statistics show significant progress, with the number of people waiting on an initial decision halving in the year through June, though the number of appeals against decisions has increased.

Accuracy Concerns and Judicial Scrutiny

Researchers and officials have expressed concern over the accuracy of the tools, particularly given the high stakes involved in determining whether individuals face persecution in their home countries. According to government estimates, using the tools together could save around an hour per case, helping decision-makers save time when reviewing interview transcripts and policy documents.

However, ORG found that the percentage of summaries deemed unfit for use due to factual errors or misrepresentations is significantly higher than the 9 percent error rate calculated in the pilot evaluation. Furthermore, a UK judge accused the Home Office of citing an AI-hallucinated document to refuse an asylum claim.

Bias, Regulation, and Transparency

Campaigners and legal researchers have also warned that well-documented cultural biases in GPT models could cause language barriers and the loss of cultural nuances, potentially leading to discriminatory decisions. Meanwhile, OpenAI has highlighted its ongoing work to improve model accuracy, alongside plans regarding the retiring of GPT-4 models.

Under the UK's AI Playbook, departments are required to maintain transparency regarding algorithmic tools used in decision-making processes and list them on a public repository. The Home Office's case summarisation and policy search tools are notably absent from that list, drawing criticism from lawmakers and privacy advocates over the absence of regulation and public disclosure.

Home Office Response

A Home Office spokesperson defended the department's procedures, stating that asylum decisions are made by trained caseworkers rather than automated systems. The spokesperson emphasized that AI tools are used strictly to assist in navigating interview records and policy information, and cannot be used in isolation to make final decisions on claims.