Digital Gatekeepers: Exploring Large Language Model's Role in Immigration Decisions

With globalization and increasing immigrant populations, many countries' immigration departments face the numerous workload with its limited staff. For instance, the Home Office Immigration and Nationality Directorate in the UK has faced increased workloads, leading to significant backlogs and administrative challenges (Yeo, 2022). Similarly, immigration judges in the USA are experiencing burnout due to enormous caseloads (Lustig et al., 2008). At the same time, these offices also face the significant challenge of ensuring fairness in their decision-making processes. Although immigration officers often view themselves as objective administrators regarding the entry and stay of immigrants (Armenta, 2012), research shows that their decisions are profoundly influenced by personal attributes (Dinesen et al., 2016), and broader social norms (Turper et al., 2015), leading to biased and discriminatory outcomes (Coates and Carr, 2005). Studies reveal that officers' decisions can be affected by emotions, stereotypes, and cultural values, resulting in profiling and differential treatment of immigrants based on nationality, race, and religion (Armenta, 2012; Dekkers, 2018).