NPR84%

Some local police have access to an ICE facial recognition app 27%

By Meg Anderson0%

6/19/2026, 2:00:00 AM

BS Summary: This article contains 28 faulty reasoning types, including Appeal to Authority, Availability Heuristic, and Anecdotal, with Negativity Bias as the most egregious example at 21.8% saturation with 231 hits. Analysis detected 1,836 faulty-reasoning hits from 1,058 analyzed words, generating a BS Score of 38.5% and a BS Rank of 27% (15,987 of 21,887 articles). This article is better (less manipulative) than 73.00% of the article peer group.

Federal immigration officers often use facial recognition technology to identify immigrants in the field. 
Now, a newly revealed document from the Department of Homeland Security outlines plans to give local police working on its behalf the same type of technology. 
The document, first reported earlier this month by the tech news outlet 404 media, is a Privacy Threshold Analysis, which is essentially a federal report assessing whether the privacy implications of a tool warrant further government study. 
The tool in question is a mobile app called the ICE Task Force Module, which allows local police to scan the faces of people they stop in their communities. 
The app then compares the facial scan against more than 250 million government records. 
Those include the State Department's Visa records and records from the Traveler Verification Service, used by the Transportation Security Administration at airports to verify identities on international flights. 
Once police scan a person's face, the app then instructs an officer either to "not detain or arrest," or it gives the officer a reference code to use to obtain more information from ICE. 
The photos captured by the app are then stored in an internal DHS system for 15 years, the document states. 
DHS declined to provide NPR with more insight about the app and how it is used. 
In a statement, the agency said ICE is committed to ensuring that the local police who partner with them have the tools needed to support ICE's mass deportation mission. 
Those local officers, called "ICE non-federal law enforcement officers" in the document, are likely participants in the federal 287(g) program. 
A subset of that program, the Task Force Model, gives local police the authority to arrest immigrants on ICE's behalf during their routine police duties. 
There are around 1,300 police agencies participating in the Task Force Model nationwide. 
The DHS analysis "raises more questions than I think it answers," says Clare Garvie, deputy director of the Technology Law and Policy Program at New York University School of Law's Policing Project. 
For one, the document says the app launched last September, which suggests police are already using it. 
It also seems to work similarly to Mobile Fortify, a facial recognition app that ICE and officers with Customs and Border Protection already use, but it's unclear if the new app uses the same technology or something entirely its own. 
Garvie says there are also questions about how and when police will deploy the app. 
"It's unclear to me whether a pre-existing stop based on some level of suspicion is required before law enforcement can use this app," Garvie says. 
"Can they walk around taking photos of whoever as sort of a dragnet way to attempt to identify individuals who might be in the country unlawfully?" 
That sort of surveillance already appears to be happening at the federal level: In places like Minnesota and Maine, community members observing ICE activity reported that federal immigration officers would take photos of their faces and license plates. 
They said the officers would often know personal information about them, including their names and where they live. 
Privacy experts told NPR that allowing local police to conduct similar surveillance could create a chilling effect on freedom of speech, if people begin to worry they'll face repercussions for attending protests, for instance, or for legally observing ICE activity in their communities. 
Homeland Security Secretary Markwayne Mullin acknowledged at a congressional hearing this month that the agency has used facial recognition technology on protesters, and had been able to identify people who were present at protests in Oregon that were also at the recent protests outside the Delaney Hall Detention Facility in Newark, N.J. 
What's more, Garvie says, facial recognition technology is not always accurate, and there have been cases of people detained by ICE who were wrongly identified by the technology. 
Patrick Eddington, a senior fellow in homeland security and civil liberties at the Cato Institute, a libertarian think tank, says giving police this capability magnifies its potential problems. 
"This kind of technology which can impact individual rights, when it's scaled, it can have potentially very, very large effects affecting lots and lots of people," he says. 
"It's like a Bill of Rights disaster pretty much waiting to happen." 
In its statement to NPR, DHS said its law enforcement methods are constitutional. 
"Like other law enforcement agencies, ICE employs various forms of technology to investigate criminal activity and support law enforcement efforts while respecting civil liberties and privacy interests," the statement said. 
But Eddington says U.S. citizens will get caught up in this surveillance. 
Officers conducting immigration enforcement, whether they are federal or local, will not know a person's citizenship status before they conduct a scan. 
"It is conceivable that a photo taken by an ICE non-federal law enforcement officer using the TFM mobile application could be that of someone other than a removable individual, including U.S. citizens," the DHS document states. 
Because every photo taken through the app is kept for 15 years, Eddington says that suggests a long-term government record of citizens and immigrants alike. 
The administration has repeatedly denied the existence of a database of protesters, despite instances in which federal agents have told community members observing them that their photo will end up in a database of "domestic terrorists." 
However, earlier this month, NPR reported on a previously unpublished letter sent to members of Congress in which former acting ICE director Todd Lyons indicates the agency gives itself wide latitude to collect information on the people its officers encounter. 
"This app wouldn't work if they didn't have databases to pull people's pictures from and compare against," says Cooper Quintin, a senior staff technologist with the Electronic Frontier Foundation, a nonprofit that advocates for digital privacy. 
"They're playing semantics. 
They're certainly not being forthright. 
You know, do they have a database of protesters? 
Maybe they don't call it that." 
He says allowing police to use this technology to do immigration enforcement is a significant expansion of ICE's operations. 
"It makes this sort of face surveillance ubiquitous on American streets," Quintin says. 
"I don't think that Americans should tolerate law enforcement being able to scan anyone's face at any time for any reason to try to determine their identity. 
This is the new form of 'papers, please.'" 
Article reasoning-pattern comparisonThis article: 7.4%Meg Anderson: 5.1%Goats and Soda (NPR): 3.4%Confirmation Bias7.4%This article: 3.5%Meg Anderson: 1.1%Goats and Soda (NPR): 1.6%Anchoring Bias3.5%This article: 12.9%Meg Anderson: 3.4%Goats and Soda (NPR): 3.8%Availability Heuristic12.9%This article: 7.0%Meg Anderson: 2.0%Goats and Soda (NPR): 1.2%Representativeness Heuristic7.0%This article: 1.6%Meg Anderson: 1.0%Goats and Soda (NPR): 0.8%Hindsight Bias1.6%This article: 0.0%Meg Anderson: 1.0%Goats and Soda (NPR): 2.4%Overconfidence Bias0.0%This article: 7.8%Meg Anderson: 10.0%Goats and Soda (NPR): 11.1%Framing Effect7.8%This article: 2.6%Meg Anderson: 1.1%Goats and Soda (NPR): 1.2%Loss Aversion2.6%This article: 2.8%Meg Anderson: 0.5%Goats and Soda (NPR): 1.1%Status Quo Bias2.8%This article: 0.0%Meg Anderson: 0.1%Goats and Soda (NPR): 0.2%Sunk Cost Effect0.0%This article: 0.0%Meg Anderson: 3.1%Goats and Soda (NPR): 3.4%Optimism Bias0.0%This article: 7.4%Meg Anderson: 3.5%Goats and Soda (NPR): 2.3%Pessimism Bias7.4%This article: 21.8%Meg Anderson: 18.0%Goats and Soda (NPR): 11.3%Negativity Bias21.8%This article: 2.7%Meg Anderson: 1.1%Goats and Soda (NPR): 2.3%Self-Serving Bias2.7%This article: 0.0%Meg Anderson: 0.2%Goats and Soda (NPR): 1.1%Fundamental Attribution Error0.0%This article: 0.0%Meg Anderson: 0.0%Goats and Soda (NPR): 0.3%Actor-Observer Bias0.0%This article: 0.0%Meg Anderson: 0.9%Goats and Soda (NPR): 2.3%In-Group Bias0.0%This article: 0.0%Meg Anderson: 2.2%Goats and Soda (NPR): 1.2%Out-Group Homogeneity Bias0.0%This article: 0.0%Meg Anderson: 0.2%Goats and Soda (NPR): 2.8%Halo Effect0.0%This article: 0.0%Meg Anderson: 0.4%Goats and Soda (NPR): 0.3%Horn Effect0.0%This article: 0.0%Meg Anderson: 0.0%Goats and Soda (NPR): 0.0%Dunning-Kruger Effect0.0%This article: 6.2%Meg Anderson: 2.7%Goats and Soda (NPR): 1.7%Recency Bias6.2%This article: 0.0%Meg Anderson: 0.5%Goats and Soda (NPR): 0.5%Primacy Effect0.0%This article: 0.0%Meg Anderson: 0.0%Goats and Soda (NPR): 0.1%Blind-Spot Bias0.0%This article: 0.8%Meg Anderson: 0.4%Goats and Soda (NPR): 1.2%Ad Hominem0.8%This article: 0.0%Meg Anderson: 0.2%Goats and Soda (NPR): 0.3%Straw Man0.0%This article: 15.4%Meg Anderson: 8.5%Goats and Soda (NPR): 6.1%Appeal to Authority15.4%This article: 3.2%Meg Anderson: 1.1%Goats and Soda (NPR): 1.5%False Dilemma3.2%This article: 5.0%Meg Anderson: 3.8%Goats and Soda (NPR): 1.3%Slippery Slope5.0%This article: 0.0%Meg Anderson: 0.8%Goats and Soda (NPR): 0.1%Circular Reasoning0.0%This article: 9.9%Meg Anderson: 7.3%Goats and Soda (NPR): 4.4%Hasty Generalization9.9%This article: 0.0%Meg Anderson: 0.0%Goats and Soda (NPR): 0.3%Red Herring0.0%This article: 0.0%Meg Anderson: 0.9%Goats and Soda (NPR): 0.7%Bandwagon0.0%This article: 11.0%Meg Anderson: 8.0%Goats and Soda (NPR): 6.6%Appeal to Emotion11.0%This article: 3.4%Meg Anderson: 1.1%Goats and Soda (NPR): 0.9%Begging the Question3.4%This article: 6.4%Meg Anderson: 4.5%Goats and Soda (NPR): 2.8%Post Hoc (False Cause)6.4%This article: 0.0%Meg Anderson: 0.0%Goats and Soda (NPR): 0.2%Tu Quoque0.0%This article: 1.7%Meg Anderson: 0.4%Goats and Soda (NPR): 0.4%Burden of Proof1.7%This article: 2.8%Meg Anderson: 0.3%Goats and Soda (NPR): 0.2%Appeal to Nature2.8%This article: 0.0%Meg Anderson: 0.2%Goats and Soda (NPR): 0.2%Composition/Division0.0%This article: 12.7%Meg Anderson: 3.6%Goats and Soda (NPR): 2.8%Anecdotal12.7%This article: 0.0%Meg Anderson: 0.0%Goats and Soda (NPR): 0.1%No True Scotsman0.0%This article: 6.9%Meg Anderson: 1.6%Goats and Soda (NPR): 1.5%Ambiguity (Equivocation)6.9%This article: 0.0%Meg Anderson: 0.2%Goats and Soda (NPR): 0.0%Gambler’s Fallacy0.0%This article: 0.0%Meg Anderson: 0.0%Goats and Soda (NPR): 0.1%Middle Ground0.0%This article: 3.0%Meg Anderson: 0.3%Goats and Soda (NPR): 0.1%Personal Incredulity3.0%This article: 0.0%Meg Anderson: 0.0%Goats and Soda (NPR): 0.2%Special Pleading0.0%This article: 0.0%Meg Anderson: 0.0%Goats and Soda (NPR): 0.5%Genetic Fallacy0.0%This article: 2.4%Meg Anderson: 0.6%Goats and Soda (NPR): 1.3%Unattributed Quote2.4%This article: 0.0%Meg Anderson: 0.9%Goats and Soda (NPR): 0.9%Quote-first Misdirection0.0%This article: 0.0%Meg Anderson: 2.3%Goats and Soda (NPR): 3.1%Biased Writer Voice0.0%This article: 2.6%Meg Anderson: 1.7%Goats and Soda (NPR): 1.0%Indoctrination2.6%This article: 2.6%Meg Anderson: 1.5%Goats and Soda (NPR): 1.1%Politically Left Leaning Bias2.6%This article: 0.0%Meg Anderson: 0.0%Goats and Soda (NPR): 0.3%Politically Right Leaning Bias0.0%This article: 0.0%Meg Anderson: 0.3%Goats and Soda (NPR): 0.3%Attempt to Sell a Product or S…0.0%

1058 words analyzed.

Speakers

5speakers49%attributed speech541writer words
Voice mapSelect a segment to jump to its words
Writer's voice • 11 words • 0.0% coverageWriter's voice • 14 words • 0.0% coverageWriter's voice • 26 words • 0.0% coverageWriter's voice • 37 words • 0.0% coverageWriter's voice • 29 words • 0.0% coverageWriter's voice • 14 words • 0.0% coverageWriter's voice • 28 words • 0.0% coverageWriter's voice • 34 words • 0.0% coverageWriter's voice • 20 words • 0.0% coverageWriter's voice • 16 words • 0.0% coverageDepartment of Homeland Security • 29 words • 0.0% coverageWriter's voice • 20 words • 0.0% coverageWriter's voice • 25 words • 0.0% coverageWriter's voice • 13 words • 0.0% coverageClare Garvie • 32 words • 0.0% coverageWriter's voice • 17 words • 0.0% coverageWriter's voice • 40 words • 0.0% coverageClare Garvie • 15 words • 0.0% coverageClare Garvie • 25 words • 100.0% coverageClare Garvie • 26 words • 0.0% coverageWriter's voice • 38 words • 0.0% coverageWriter's voice • 18 words • 0.0% coverageWriter's voice • 43 words • 0.0% coverageMarkwayne Mullin • 52 words • 0.0% coverageClare Garvie • 28 words • 0.0% coveragePatrick Eddington • 28 words • 100.0% coveragePatrick Eddington • 28 words • 0.0% coveragePatrick Eddington • 12 words • 0.0% coverageDepartment of Homeland Security • 13 words • 0.0% coverageDepartment of Homeland Security • 30 words • 0.0% coveragePatrick Eddington • 12 words • 0.0% coverageWriter's voice • 22 words • 0.0% coverageDepartment of Homeland Security • 36 words • 0.0% coveragePatrick Eddington • 25 words • 0.0% coverageWriter's voice • 36 words • 0.0% coverageWriter's voice • 40 words • 0.0% coverageCooper Quintin • 36 words • 0.0% coverageCooper Quintin • 3 words • 0.0% coverageCooper Quintin • 5 words • 0.0% coverageCooper Quintin • 9 words • 0.0% coverageCooper Quintin • 6 words • 0.0% coverageCooper Quintin • 19 words • 0.0% coverageCooper Quintin • 13 words • 0.0% coverageCooper Quintin • 27 words • 100.0% coverageCooper Quintin • 8 words • 0.0% coverage
Selected voice

Patrick Eddington

100%flagged-word coverage
105 attributed words20% of attributed speech80% writer coverage
0%15.0%30.0%Politically Left Leaning B+26.7 ptsWriter: 0.0%Patrick Eddington: 26.7%26.7%

Attribution is sentence-level. Pattern percentages are calculated only from words assigned to that voice.

Loading…
Loading…
Loading…
Loading…

Analysis

Hover over highlighted words in the article to view the associated bias or fallacy analysis.