At a Glance
- An AI tool misread résumés and funneled non-cops into a four-week online course instead of the required eight-week academy
- ICE must now recall an unknown number of recruits to the Federal Law Enforcement Training Center in Georgia
- The error occurred during a rushed 10,000-officer hiring surge backed by $50,000 signing bonuses
- Why it matters: Immigration agents with less training are operating in U.S. cities as arrests surge and local backlash grows
ICE’s push to add 10,000 new officers by the end of 2025 hit a major snag when artificial intelligence software misclassified applicants, sending people without law-enforcement backgrounds into a shortened training track meant only for seasoned cops, according to two officials with direct knowledge of the mistake.
The AI system scanned résumés for the word “officer” and automatically routed those applicants into ICE’s LEO program, a four-week online course designed for recruits who already carry a badge. Titles such as “compliance officer” or statements like “aspiring ICE officer” triggered the shortcut, the officials said.
How the AI Error Unfolded
- The tool was deployed as ICE raced to meet a White House mandate for 10,000 new hires
- Résumés containing “officer” were tagged as experienced law-enforcement personnel
- Applicants were placed in the four-week remote course instead of the eight-week, in-person academy
- The mistake went unnoticed for more than a month, until mid-fall
“They were using AI to scan résumés and found out a bunch of the people who were LEOs weren’t LEOs,” one official told News Of Losangeles.
The regular academy curriculum in Georgia covers immigration law, firearms handling, and physical-fitness tests. Recruits who bypassed it missed those core modules.
Recall and Retraining
Once the error surfaced, ICE halted the AI filtering and began manual reviews of every recent hire, the officials said. The agency now faces the logistical headache of bringing improperly trained recruits back to the Federal Law Enforcement Training Center.
“They now have to bring them back to FLETC,” the second official confirmed.
Neither official knew the exact number of officers affected or how many may already have begun making immigration arrests in the field. Both spoke on condition of anonymity because they were not authorized to discuss the internal matter publicly.
The Department of Homeland Security did not respond to a request for comment.
Hiring Surge Under Scrutiny
Congress allocated funding under the One Big Beautiful Bill, allowing ICE to offer $50,000 signing bonuses to attract recruits. While the agency technically met its 10,000-officer target on paper, the AI mix-up means the true number of fully trained officers on the street is lower.
Field offices provide supplemental training before officers conduct solo operations, and the officials emphasized that recruits flagged by the AI likely received that extra instruction. Still, the shortcut undercuts the rigorous standards ICE set for its expansion.
Broader Enforcement Climate
The revelation lands amid rising tension over ICE tactics. Local police departments, community groups, and lawmakers have questioned the agency’s approach following the fatal shooting of Renee Nicole Good in Minneapolis by veteran ICE officer Jonathan Ross. Ross had more than a decade of experience and would not have been routed through the AI screening for new hires.
Since Nov. 29, over 2,400 people have been arrested in Minneapolis alone, after ICE deployed more than 2,000 officers to the area, according to DHS spokesperson Tricia McLaughlin. Minnesota has filed suit seeking to remove DHS from the state.
Next Steps
ICE has suspended the AI tool and reverted to human review of all incoming résumés. The agency must also decide whether to pause deployments until every misclassified recruit completes the full academy regimen.
Key takeaways:
- An AI résumé scanner mistook job titles for police experience
- Untrained officers entered field offices under a fast-track program
- ICE is recalling staff to finish mandatory academy training
- The glitch complicates the agency’s goal of 10,000 fully qualified officers by the end of 2025
