AI Error Deployed Untrained ICE Agents, Sparking Security Concerns
AI glitch sent untrained ICE agents into field

A significant error in an artificial intelligence system led to hundreds of newly recruited Immigration and Customs Enforcement (ICE) agents being deployed to field offices without completing mandatory training, a new report has revealed.

How the AI Recruitment Tool Failed

According to sources speaking to NBC News, the incident occurred in mid-fall 2025 during a major recruitment drive. ICE employed an AI tool to scan resumes for its "LEO program," designed for candidates who are already certified law enforcement officers. This program requires only four weeks of online training.

However, applicants without any prior police or federal law enforcement background are supposed to undergo a much more rigorous eight-week in-person course at the Federal Law Enforcement Training Center (FLETC) in Georgia.

The AI's critical mistake was to sort all applicants with the word "officer" in their resume into the shorter training stream. This included individuals who had worked as different types of officers, such as safety or administrative officers, and even those who simply expressed a hope to one day become an ICE officer.

Consequences and Corrective Actions

The result was that a large number of the new recruits were incorrectly flagged as experienced law enforcement officers. Officials stated that most of the new applicants were mis-categorised, despite many having no relevant experience.

"They were using AI to scan resumes and found out a bunch of the people who were LEOs weren't LEOs," one official said. The exact number of officers who received improper training remains unclear, as does how many may have already begun making immigration arrests.

Upon discovering the error, ICE immediately began manually reviewing the resumes of new hires. "They now have to bring them back to FLETC," an official confirmed, indicating a costly and disruptive retraining process.

Broader Context and Political Fallout

The recruitment surge and this AI mishap come amid a significant increase in ICE's operational presence across the United States. The agency's activities have sparked intense political debate and public protest, particularly in states like Minnesota.

In Minnesota, where over 2,000 immigration arrests have been reported since enforcement operations escalated in December, widespread protests have erupted. The situation grew more tense following the shooting death of Renee Nicole Good in Minneapolis by an experienced ICE officer, an incident unrelated to the training error.

On Thursday 15 January 2026, former President Donald Trump responded to the protests on his Truth Social platform. He labelled protesters as "professional agitators and insurrectionists" and threatened to invoke the rarely used Insurrection Act of 1807 to deploy military troops if "corrupt politicians" in Minnesota did not stop the attacks on ICE agents.

The Department of Homeland Security has not yet provided an official comment on the AI training failure. Officials noted that field offices do provide some additional training, but the core academy instruction was missed by those affected by the algorithmic error.