The Chartered Institute of Linguists and Institute of Translation and Interpreting have written to the UK Home Office flagging concerns regarding the Home Office’s Auto-Translation and Transcription RFI, released last week, and urging engagement with researchers from the University of Bristol, CenTraS-UCL and CTS Surrey on the growing research base on the risks of AI use for translation in public service contexts.

As the UK’s largest professional bodies for translators, we encourage UK Government to actively engage with the growing range of academic research on the use of AI in public services contexts, all of which points out major issues with accuracy and reliability – especially when used in high stakes situations and for lower resource languages, which are commonly those in demand in the immigration and asylum system.

Put simply, Asylum & Human Rights Operations, Border Force, Immigration Enforcement, Visas, Status & Immigration (VSI) and Passports, Citizenship & Civil Registration are all contexts where the red flags which this research identifies are very real, and present significant risks to human rights and public safety.

We have strongly advised that alongside seeking input from suppliers interested in taking part in this RFI, the Home Office also connect with the research and researchers from Bristol, UCL and Surrey whose work is summarised and can be accessed here: Uses of AI Translation in UK Public Service Contexts | CIOL (Chartered Institute of Linguists)

AI is powerful and useful in many contexts, but drawing on the available research will enable Government to better navigate the balance of risk in using these technologies in the potentially high-stakes contexts the Home Office RFI references.