An AI blunder reportedly sent unprepared federal immigration agents into the streets.
In a stunning tech failure, U.S. Immigration and Customs Enforcement's rush to beef up its force with 10,000 new officers blew up in its face when a faulty artificial intelligence system sent undertrained recruits into the field, two law enforcement officials familiar with the error told NBC News.
The bungled AI tool was supposed to surface applicants with law enforcement credentials for ICE's "LEO program," a four-week online training shortcut for experienced law enforcement officers. Instead, the system went haywire, flagging anyone with "officer" in their résumé — including compliance officers and aspiring agents — for the abbreviated course. Applicants with no law enforcement background have to take an eight-week training course in Georgia, where they learn how to handle a gun and undergo fitness tests.
“They were using AI to scan résumés and found out a bunch of the people who were LEOs weren’t LEOs,” one official said.
Most new hires were wrongly classified as law enforcement veterans, even though many had zero experience with police or a federal agency.
The blunder wasn't caught until mid-fall, over a month into the recruitment blitz. ICE scrambled to implement manual reviews and order the mislabeled agents back to proper training.
The $50,000 signing bonus hiring spree, funded by Congress's One Big Beautiful Bill, officially hit its 10,000-officer target on paper. But behind the numbers, the agency's actual street-ready force fell far short, the officials admitted, as improperly trained agents required remedial courses before deployment.