The Department of Homeland Security (DHS) is planning to initiate artificial intelligence (AI) testing through three pilot projects including identifying fentanyl and addressing child sexual exploitation. Additionally, tests will be conducted in collaboration with the Federal Emergency Management Agency (FEMA), United States Citizenship and Immigration Services (USCIS).
"The use of AI could improve the efficiency and effectiveness of USCIS adjudications, but USCIS's track record is always toward layering on more reviews, not streamlining the existing process, so I'm skeptical that this will do anything but throw up more spurious red flags for adjudicators to have to manually sort through," said David J. Bier, associate director of immigration studies at the Cato Institute.
FEMA plans to utilize artificial intelligence (AI) to assist communities in strategizing and crafting mitigation plans to bolster resilience and reduce potential risks. Meanwhile, USCIS intends to employ AI to enhance the training of immigration officers.
"The unprecedented speed and potential of AI's development and adoption presents both enormous opportunities to advance our mission and risks we must mitigate," said DHS Secretary Alejandro Mayorkas in a Monday announcement. In addition to introducing three pilot projects, the Department of Homeland Security (DHS) unveiled a set of guiding principles in a roadmap to steer its implementation of AI initiatives.
"The use of AI to interdict narcotics and to combat child sex trafficking has merit, but it should used as a tool to augment efforts to secure our borders and prevent the cartels from bringing drugs into the country and exploiting children, said Ira Mehlman, media director at The Federation for American Immigration Reform. "It is not a substitute for real border enforcement," he added.
The plan outlines that the Department of Homeland Security (DHS) must guarantee that its implementation of AI prioritizes the full respect of privacy and civil rights. It emphasizes thorough testing to prevent bias, privacy violations, and other potential risks. Additionally, the roadmap encourages the agency to advance national AI safety and security efforts, such as safeguarding American cyber networks and critical infrastructure.
"In conjunction with policies that deter mass illegal immigration – such as the reinstatement of Remain in Mexico – and preventing large-scale abuse of our asylum system, AI offers another layer of protection and deterrence, but it would have limited effect so long as the current chaos prevails, Mehlman told The Foreign Desk. "AI alone cannot compensate for CBP's manpower being tied up apprehending, processing, and releasing millions of illegal aliens each year," he added.
The plan outlines that the Department of Homeland Security (DHS) will lead in forming robust partnerships with the private sector, local governments, and research institutions. The AI strategy is an extension of DHS's recent announcement to recruit 50 experts in AI technology. The department plans to employ AI models such as OpenAI's ChatGPT to aid in implementing the new tests. Additionally, it has forged partnerships with Anthropic, Meta, Microsoft, Google, and Amazon as it initiates the pilot programs.
Related Story: California Seized Enough Fentanyl Last Year to Kill Entire World ‘Nearly Twice Over’
