ICE street arrests surge across the United States following the launch of the Trump administration, according to new federal and academic reports analyzing immigration enforcement trends.
Data released in the Department of Homeland Security’s (DHS) annual performance report and UCLA’s Deportation Data Project show that Immigration and Customs Enforcement (ICE) has dramatically expanded arrests, surveillance capacity, and interagency cooperation in recent years.
The reports suggest that enforcement activity has intensified not only through technology and data-sharing agreements but also through a sharp rise in arrests conducted in public spaces such as streets, workplaces, homes, and immigration courts.
Enforcement numbers far exceed federal targets
According to the DHS report, ICE arrested 167,651 noncitizens with criminal histories or pending criminal cases during fiscal year 2025.
The figure is more than double the agency’s annual arrest target of 75,000 and represents a 106% increase from the previous year.
Following the surge in enforcement activity, ICE has raised its future annual arrest target to 400,000 cases, signaling a significant expansion of immigration enforcement priorities.
DHS officials attributed the increase to stronger interagency coordination, improved case identification tools, and greater use of advanced technologies.
“Enhanced collaboration with partner agencies, improved data analytics, and targeted enforcement against public safety threats contributed to the increase in arrests,” DHS said in the report.
AI tools and surveillance expand enforcement reach
Federal officials say ICE has increasingly relied on artificial intelligence to support investigations and enforcement operations.
Agents now use AI systems to process tips, analyze investigative data, and verify identities during enforcement actions.
Currently, ICE operates about 24 AI-based systems, including eight core platforms that directly influence major decisions such as arrests or restrictions on individual rights.
However, the expansion of automated analysis has also raised concerns among legal experts.
Immigration attorney Brian Oh warned that while AI and data analytics can help locate undocumented migrants who have moved residences, they may also create incentives for aggressive enforcement.
“The incentive structure could lead to overly aggressive tracking efforts aimed at boosting enforcement statistics,” Oh said.
Beyond AI systems, DHS has also expanded data collection and information-sharing networks across federal agencies.
ICE now receives information from agencies such as the Internal Revenue Service (IRS), international law enforcement networks, and the Centers for Medicare & Medicaid Services (CMS).
ICE street arrests surge nationwide, UCLA report finds
Independent research also confirms the rapid escalation in enforcement.
A UCLA Deportation Data Project report found that deportations increased fivefold during the first year of the current administration, while immigration enforcement arrests nationwide rose more than four times compared with the final period of the previous administration.
The most striking trend was the sharp increase in public or street-based enforcement.
Researchers found that monthly street arrests rose from roughly 400 in January 2025 to nearly 4,500 by January 2026, representing more than an 11-fold increase.
Arrests are now increasingly occurring during routine activities, including at homes, workplaces, immigration court appearances, and mandatory check-ins.
Detention levels have also climbed sharply. Average daily immigration detention rose from about 14,000 people in 2024 to approximately 57,000 by January 2026, according to the study.
Immigration attorneys warn that enforcement is becoming increasingly embedded in everyday life.
“ICE was not originally designed to intervene in daily settings like this,” said Andrew Nieter, an attorney with the American Immigration Lawyers Association (AILA).
“With enforcement expanding to courts, homes, workplaces, and even airports, immigration policing is becoming normalized in everyday environments.”
