ICE Buys 6,000 Pairs of Electric Shock Gloves

Hands in handcuffs gripping jail bars — Photo: Anelo / Shutterstock

An Army study that impressed ICE enough to justify a $16.7 million purchase order also contains the words ICE would rather not quote: the device was never tested for safety, and the manufacturer’s own manual warns against most of the uses it may now get.

Key Points

A previously unreleased Army study from Fort Leavenworth found that electric shock gloves stopped resistance within seconds, prompting ICE to order 6,000 pairs under a $16.7 million contract.
The gloves work by overloading peripheral nerve endings in the skin — producing pain rather than the muscle lockup caused by a Taser.
Watchdog group American Oversight says the same Army report measured “compliance time” but did not evaluate safety, and never assessed risk to children, pregnant women, or people with disabilities.
The manufacturer’s own product manual warns against using the gloves as punishment, against people who are merely verbally defiant, or on high-risk populations — and notes they may contribute to death or serious injury.
The gap between a controlled military exercise and real-world ICE enforcement — arrests, protests, crowded transport vans — remains untested and largely unaddressed by public policy.

How the Device Actually Works

The gloves at the center of this controversy — officially the Generated Low Output Voltage Emitter, or GLOVE, built by Compliant Technologies — are not Tasers, and the distinction matters more than it sounds. A Taser’s prongs disrupt motor neurons, causing involuntary muscle contraction that effectively freezes a target in place. The GLOVE instead must touch bare skin directly, delivering a voltage that, according to the Army’s own study, “works by overloading the peripheral nerve endings” to produce pain rather than paralysis. The compliance it produces is psychological and sensory, not mechanical: the subject stops resisting because the pain is unbearable, not because the body has lost the ability to move.

That mechanism is precisely why the device reads so differently depending on which side of the debate is describing it. ICE and the manufacturer frame it as a “de-escalation” tool positioned between bare-hand restraint and higher-force options like batons or firearms. Critics counter that a device whose entire function is to inflict pain until someone submits is, definitionally, a compliance weapon — not a buffer against force, but a more concealable form of it.

What the Army Actually Found

In October 2024, the Army ran a previously unreleased exercise at the military prison in Fort Leavenworth, Kansas, using active-duty soldiers in roleplay scenarios meant to mirror real detention work: forcibly removing a combative prisoner from a cell, breaking up a disturbance in a common area, and moving an uncooperative inmate during transport. The results, as reported by the Associated Press and NBC News, were dramatic — participants “felt intense pain, stopped fighting immediately and feared additional exposures,” and the devices shortened the length of physical encounters enough that the Army report described measurable gains in team efficiency. Soldiers were left with fatigue, soreness, and what the study itself called “excessive apprehensiveness to being shocked again”.

ICE’s reaction to that data was not caution — it was procurement. According to NBC’s reporting, the agency signed a $16.7 million contract for 6,000 pairs of the gloves after the study’s findings were already known. That sequencing matters: this was not a pilot program awaiting further evidence. It was a scaled operational commitment made with the pain-and-fear findings already on the record, which is the detail watchdog groups have seized on hardest.

The Safety Question the Study Never Answered

The most consequential fact in this entire record may be the one buried in a FOIA release rather than a headline. American Oversight, which obtained the underlying Army memo, summarized its core finding bluntly: the report showed the gloves “force compliance through pain” and “did not evaluate safety”. The watchdog further noted the assessment said nothing about risk to children, pregnant women, or people with disabilities — populations that, in the ordinary course of immigration enforcement, ICE agents do in fact encounter.

That gap is not a rhetorical flourish; it is confirmed independently by the manufacturer’s own documentation. PBS NewsHour reported that Compliant Technologies’ product manual explicitly warns the device should not be used as punishment, against someone exhibiting only “verbal defiance or belligerence,” or on children, pregnant women, or elderly or disabled people. The same manual, per PBS, states the gloves “may contribute to death or serious injury” and that the shock “can increase blood pressure, heart rate and adrenaline and can cause changes in heart rhythm”. A device whose own warning label rules out exactly the scenarios — passive non-compliance, mixed-population encounters — that define much of ICE’s daily street work is not a hypothetical liability. It is a documented one, stated by the company that built it.

A Controlled Exercise Is Not a Street Encounter

Even setting the safety gap aside, the Fort Leavenworth test measured something narrower than ICE’s marketing implies. Active-duty soldiers role-playing a scripted prison scenario, evaluated on “compliance time” — how long it took officers to restrain the subject — is a controlled, consenting, institutional environment. It tells you little about how the device performs during a chaotic arrest on a sidewalk, a crowded transport van, or a protest, where subjects are not soldiers following a training script and where supervisory oversight is thinner. The National Institute of Justice has long flagged this exact transferability problem with less-lethal technology generally, commissioning research specifically into in-custody deaths tied to conducted-energy devices because lab conditions routinely fail to predict field outcomes.

White House border czar Tom Homan has defended the purchase publicly, describing the gloves as a way to end confrontations without resorting to deadly force. That is a coherent policy argument, and it is not without foundation — the Army data does show the device working quickly in the scenario it was tested in. But “it worked in a scripted prison drill” and “it is safe and appropriate for ICE’s actual operating environment” are two different claims, and only the first one has real evidence behind it. Congressional correspondence, including a letter from Rep. Delia Ramirez to Rep. Markwayne Mullin, has already flagged the gloves as “designed to inflict severe pain without leaving physical evidence or visible optics” — a combination that invites abuse precisely because it resists documentation.

ICE has awarded a $16.7 million no-bid contract for 6K pairs of electric-shock gloves manufactured by Kentucky-based Compliant Technologies. And they're calling it "de-escalation."

Now a previously unreleased US Army study, obtained through a FOIA lawsuit by American Oversight,…

— IncarcerNation.com (@IncarcerNation) October 8, 2026

What Happens Next

The honest accounting here is not that the Army study was fabricated or that ICE’s efficacy claim is baseless — the device plainly did what it was designed to do in the test conditions where it was tried. The honest accounting is that efficacy and safety are separate questions, and only one of them has been answered. Until the full report, ICE’s internal use-of-force guidance, and independent field data are made public, the public is left weighing a documented capability against a documented warning label — and a procurement decision that moved forward without closing that gap.