BLACK mobile logo

international

Decolonizing AI at the U.S. border

April 25, 2026

A report by the Black Alliance for Just Immigration and UC Irvine law clinics examines how artificial intelligence systems used in U.S. border enforcement and immigration control disproportionately harm Black migrants and migrants of color through algorithmic bias. The study argues that AI technologies employed throughout the immigration process—from surveillance towers and drones at borders to facial recognition apps and risk assessment algorithms—systematically discriminate against darker-skinned individuals and non-English speakers while violating international anti-discrimination treaties the U.S. has ratified. The organizations advocate for a "decolonial approach" to AI development that centers African philosophical frameworks and ensures affected communities participate in designing these systems.

Who is affected

  • Black migrants and migrants of color attempting to enter or remain in the United States
  • Asylum seekers fleeing violence, persecution, and torture
  • Nigerian nationals (specifically mentioned as disproportionately flagged after 2020 travel restrictions)
  • Non-English speaking applicants, particularly those speaking niche languages
  • Migrants with darker skin tones (misidentified by facial recognition at rates 10-100 times higher than white faces)
  • Individuals detained or under surveillance by ICE
  • Migrants with atypical documentation

What action is being taken

  • Customs and Border Patrol (CBP) is deploying autonomous surveillance towers and drones to identify "objects of interest" at the U.S.-Mexico border
  • The Automated Targeting System (ATS) is compiling domestic and international databases to predict visa overstays
  • Immigration and Customs Enforcement (ICE) is using predictive algorithms including "Hurricane Score" for surveillance determination
  • ICE is using the RAVEn platform to analyze trends across data sources from 56 countries
  • U.S. Citizenship and Immigration Services (USCIS) is using Asylum Text Analytics (ATA) to detect fraud in applications
  • USCIS is using an AI-powered Evidence Classifier to review documentation for adjudicators
  • Debates about reinstating the CBP One app are continuing under the current administration

Why it matters

  • This matters because the U.S. is legally obligated under the International Convention on the Elimination of All Forms of Racial Discrimination (ICERD), which it ratified in 1994, to prevent racial discrimination and ensure equal treatment before the law. The deployment of biased AI systems in immigration enforcement violates these international commitments while producing tangible harmful outcomes, including historically high migrant death rates as surveillance pushes people into more dangerous terrain. The lack of transparency, absence of consent mechanisms, and denial of remedies for those harmed by algorithmic decisions undermines fundamental human rights protections while perpetuating structural racial discrimination under the guise of technological efficiency and border security.

What's next

  • Federal laws governing DHS's use of AI must prohibit racially discriminatory results and mandate discrimination-prevention measures, independent oversight, public disclosures, stakeholder consultation with diverse populations, and access to effective remedies
  • African and diaspora communities must be actively involved in conceptualizing, inventing, innovating, and operating AI systems
  • Individuals negatively affected by AI must be promptly notified and given the option to opt out when appropriate
  • City policies must include explicit pledges not to share information with DHS for AI development or deployment
  • Until AI systems are free of discrimination and diverse perspectives are meaningfully included, they must not be allowed to be used on any border

Read full article from source: Global Voices