The Open Rights Group (ORG) says 45 of 203 AI-generated interview summaries failed initial expert checks after the system was rolled out in April this year, a failure rate of 22 percent compared with the 9 percent rejection rate recorded during the original pilot. Another 19 summaries were flagged for further review, and 30 of 56 second-stage reviews failed to meet the required standard.
The findings, based on Freedom of Information requests and correspondence with ministers, concern two tools used by Home Office caseworkers: Asylum Case Summarisation (ACS), which summarises asylum interviews, and Asylum Policy Search (APS), a chat-style tool for searching country policy information. Both run on OpenAI's GPT-4o model through Microsoft's Azure AI Foundry.
Errors can include factual inaccuracies and misrepresentations of what asylum seekers said during interviews. ORG also found that summaries could pass checks even if they omitted important details, as long as the information they did include was accurate.
Despite the results, few summaries have been examined. ORG estimates Home Office experts checked only 0.3 percent of them between April and July, based on the department's projected monthly caseload of 17,515.
The group also questions whether the tools are saving much time. The average monthly number of initial substantive asylum decisions since April fell about 42 percent compared with the same period in 2025, while the mean time from interview to decision rose 29 percent. Those figures do not prove AI caused the slowdown, ORG notes, but they offer little evidence of the productivity gains ministers have promised.
Oversight of the separate Asylum Policy Search tool also appears lacking. The Home Office confirmed in August that expert testing had stopped, despite previously telling MPs that checks were continuing. ORG says no summary of the testing results was produced.
Asylum applicants are not specifically told when AI has been used in their cases and cannot inspect or correct the summaries generated from their interviews. The department's own data protection assessment says ACS automates work previously done by technical specialists, who reviewed and summarised cases before passing them to decision-makers. ORG warns the tool could remove an experienced layer of scrutiny rather than simply save staff time.
The government insists humans remain responsible for asylum decisions, but ORG questions how reliably caseworkers can spot mistakes in AI-generated information. It is calling for both tools to be suspended and independently audited.
The Home Office did not respond to questions from The Register. The department's technology plans have faced criticism before: in June, more than 60 rights groups urged ministers to abandon plans for AI-powered age estimation of asylum-seeking children, and an inspection found the long-delayed Atlas immigration system was failing to provide caseworkers with feedback from asylum appeals.