The United States Department of Homeland Security (DHS) is now looking at artificial intelligence to act like refugees during practice interviews for immigration officers, as confirmed by Secretary Alejandro Mayorkas.
A bid to help train the officers in making more accurate decisions while reviewing applicants for refugee status.
According to Mayorkas, the effort tackles labor-intensive training that usually entails senior staff members. He stated that DHS is training robots to simulate refugees in this pilot program so that officers may get experience interrogating them.
In particular, the department announced that United States Citizenship and Immigration Services, an organization under DHS, would develop an AI program that would customize training materials to officers' requirements and equip them to make more accurate decisions.
According to sources, AI will not decide on immigration independently; however, according to Mayorkas, AI will assist officers by identifying conditions unique to each country and other relevant data.
Read Also: US Police Banned by Microsoft from Using Azure's AI Facial Recognition Technology
DHS's AI Initiatives
Speaking outside the security-focused RSA Conference in San Francisco, the speaker expanded on DHS's earlier-in-the-year announcements about artificial intelligence (AI).
According to the government, it intends to use so-called generative AI, which generates original information from historical data, to build an interactive app to enhance the training it provides to immigration agents.
The pilot is an addition to the numerous experiments conducted by the government and industry to use AI to lower costs and boost performance, especially after ChatGPT's 2022 viral launch. Problems with translation, wrong timeframes, and pronouns have arisen during this experimentation.
Mayorkas stated that the department has tried to identify irregularities when commercial trucks and passenger cars cross borders, among other more "advanced" uses of AI.
He stated that the intention is to assist the department in identifying attempts to smuggle fentanyl and other illegal goods into the country.
DHS' Collaborations With AI Giants
The DHS's AI initiatives align with previous reports from March, in which the organization stated that it would collaborate with Anthropic, Meta, and OpenAI on its three recently created AI-focused projects.
At the time, generative AI was reported to be implemented by the Federal Emergency Management Agency (FEMA) as part of an AI pilot program to expedite local governments' risk reduction planning processes.
Homeland Security Investigations (HSI), a branch of Immigration and Customs Enforcement (ICE), will use massive language models to quickly sift through massive data warehouses and accelerate its investigative reports. HSI investigates child exploitation, human trafficking, and drug trafficking.
Sources claim that if AI is applied to reviewing homeland security investigations involving blue pickup trucks, it might make it simpler for detectives to pinpoint suspects who fit the description.
Additionally, US Citizenship and Immigration Services (USCIS), which conducts initial assessments for asylum seekers, will train staff members using chatbots.
The organization selected Meta, Anthropic, and OpenAI to test several technologies. Microsoft, Google, and Amazon cloud service providers will be used in its trial programs.
Related Article: GPT-Based AI Chatbot for Spies: Did Microsoft Develop One for US Intelligence?
(Photo: Tech Times)
