The Trump administration and immigration enforcement agencies are reportedly expanding their use of artificial intelligence tools to help identify and locate people without legal status in the United States.
According to procurement records reviewed by The Lever, Immigration and Customs Enforcement recently approved a $12.2 million contract for an AI system that claims it can analyze behavioral patterns, daily movements and location data to help flag potential enforcement targets.
In official documents, the system is referred to as Project SAFEHAVEN and is described as an effort to reduce reliance on multiple disconnected databases and analytical tools in favor of a centralized platform designed to improve efficiency and lower costs.
The Lever reported that the system operates through a question-based interface and uses what it describes as persistent background data collection to identify patterns in activity among individuals who may be subject to immigration enforcement.
According to the procurement documents, the tool is also designed to generate real-time location analysis and to assess whether individuals or groups may be linked to alleged criminal organizations, including gangs or cartels. It can reportedly build "target profiles" by aggregating data from sources such as Wi-Fi connections and mobile devices, including smartphones and smartwatches.
The contract was awarded for use by the Homeland Security Task Force National Coordination Center, which officials describe as a hub for regional homeland security task forces.
Although originally intended to support efforts against drug trafficking and other organized crime, The Lever reported the system could also be used to identify "extremists" and "illegal re-entrants," among other categories.
The contract runs for one year, from May 1 through April 30, 2027. Work will primarily be conducted at contractor facilities, with periodic visits to ICE headquarters in the Washington metropolitan area.
According to the documents, the system can also sort data by location and categorize individuals using attributes such as gender, country of origin and other identifiers.
The procurement records state the system must comply with the U.S. Constitution and all applicable laws and regulations, including privacy and civil rights protections. They also require safeguards such as human oversight of AI-assisted decisions, documentation of AI-generated outputs and compliance with federal procedures for notice and appeal in cases involving individual determinations.
The documents further state that AI tools cannot be used as the sole basis for enforcement actions or for predictive judgments about future behavior or emotional state if such use would result in discriminatory or unlawful outcomes.
Edge Ops LLC, the company developing the system, did not respond to requests for comment from The Lever. ICE also did not respond.
© 2025 Latin Times. All rights reserved. Do not reproduce without permission.
