The United States is embarking on a significant endeavor, aiming to utilize artificial intelligence (AI) for training immigration officers responsible for evaluating refugee status applications. Alejandro Mayorkas, the Secretary of Homeland Security, disclosed to reporters on Tuesday (May 7) that the Department of Homeland Security (DHS) is initiating this AI program to tackle what he termed as “labor-intensive” tasks.
Mayorkas highlighted that the AI initiative aims to streamline the “labor-intensive” procedures usually requiring the involvement of senior personnel. The DHS’s pilot project involves training machines to mimic refugees, enabling officers to practice interviewing vulnerable individuals seeking refuge in the United States.
Addressing the security-focused RSA Conference in San Francisco, Mayorkas emphasized the challenge posed by trauma, making refugee applicants hesitant to openly discuss their experiences. He stated, “Refugee applicants, given the trauma that they have endured, are reticent to be forthcoming in describing that trauma.” Hence, the AI is being trained to exhibit similar reticence and simulate characteristics of applicants.
Regarding the functioning of the AI trainer, the DHS plans to develop an interactive application to complement the training of immigration officers. This application will utilize generative AI, which generates novel content based on historical data, to create customized training materials for officers. The DHS anticipates that this approach will enhance officers’ ability to make accurate decisions.
The development of this application falls under the purview of the DHS’s United States Citizenship and Immigration Services. However, it’s important to clarify that the AI will not autonomously make immigration decisions. Mayorkas affirmed to Reuters that the AI will possess knowledge of country-specific conditions and other relevant information to assist officers in their decision-making process.