ICE plans to give local police facial recognition app for immigration enforcement

An Immigration and Customs Enforcement (ICE) privacy document publicly accessible on the agency’s website says a Customs and Border Protection (CBP) mobile facial recognition app will allow ICE-deputized state and local law enforcement officers to photograph people in the field, run their faces against federal identity records, and help determine whether they are removable from the U.S.
The 10-page developerPrivacyPolicy (PTA), labeled ICE Task Force Module (TFM), Version ICE Privacy 09-2025, lists the app’s launch date as September 24, 2025, and says the project was in development when the document was approved by the ICE Privacy Office the next day.
The document describes the TFM as a mobile application that was developed by CBP and made available to authorized “ICE non-federal law enforcement officers” operating in the field to verify a person’s identity and immigration status through facial recognition.
The app is believed to be Mobile Fortify, a mobile biometric application already deployed by ICE and CBP which CBP released last year for local law enforcement agencies participating in ICE’s 287(g) Task Force Model.
Discovery of the TFM document, labeled “developerPrivacyPolicy.pdf” on ICE’s website, adds a new dimension to what is known about the Department of Homeland Security’s (DHS) plan to extend the app to state and local police officers working under ICE authority.
According to the document, the TFM app would run face photos against more than 250 million DHS and Department of State visa records included in subsets of CBP’s Traveler Verification Service.
The response would consist of text messages instructing the officer either not to detain or arrest the person under ICE jurisdiction, or providing a reference code the officer could use outside the app to obtain additional information from ICE.
A police officer or sheriff’s deputy acting under ICE authority could take a person’s photograph in the field, submit it through the mobile app, and receive a response that helps determine whether the encounter becomes an immigration enforcement action.
The tool could be made available to more than 1,000 local and regional law enforcement agencies. If deployed at that scale, the app would put mobile facial recognition inside the expanding network of local police departments and sheriff’s offices participating in ICE’s 287(g) program.
The 287(g) program is the section of immigration law that authorizes state and local officers to perform certain immigration officer functions under ICE direction and supervision.
287(g) agreements had expanded across 40 states by mid-2025. And by early this year, ICE had more than 1,300 agreements across more than 1,100 agencies, with the Task Force Model representing a major share of the expansion.
The proposal marks a major step in the Trump administration’s effort to push immigration enforcement deeper into state and local policing.
Rather than limiting biometric searches to federal agents, ports of entry, detention settings, or specialized investigative units, the plan would put face-scanning capability into the hands of local police officers who encounter people during routine patrols, traffic stops, jail bookings, street operations, and immigration task force work.
The document is explicit about what the app collects. It says the only data directly submitted through the mobile application are face images and associated metadata, including geolocation.
When an authorized ICE non-federal law enforcement officer takes a photograph, the geolocation is tagged to the biometric so ICE can identify where the encounter occurred.
The Privacy Threshold Analysis says photographs and fingerprints are used to verify identity against the “targeting hotlist,” language that raises additional questions because the TFM-specific section says the only information directly submitted through the app is a face image and geolocation metadata.
The storage provisions are also important. The document says the information is not stored locally on the mobile device, but in DHS’s Automated Targeting System (ATS). It says each new photograph, whether it matches a record or not, is stored and retained in ATS for 15 years.
The privacy analysis also says individuals are not provided a Privacy Act statement, privacy notice, or other notice at the time of collection.
It says the mobile application does not provide affirmative express consent before DHS uses device functions or accesses sensitive content, and it says the app does not have an app-specific privacy policy available to users in an app store or within the app.
In other words, a person scanned by a local officer using the app would not necessarily know that a face image was being submitted to DHS, that geolocation metadata were being attached to the encounter, that the image was being checked against TVS, or that the photograph could be retained in ATS for years.
The ICE Privacy Office acknowledged the sensitivity of the system, pointing out that the TFM mobile application and the technology described in the PTA are “privacy sensitive,” and that ICE non-federal law enforcement officers may use TFM to collect information in identifiable form about individuals regardless of citizenship or immigration status.
It also says it is conceivable that a photo taken through the app could be of someone other than a removable individual, including U.S. citizens.
The document further says ICE non-federal law enforcement officers do not know an individual’s citizenship when first encountered and would use the TFM app to determine or verify identity and confirm whether the person is a match to CBP TVS.
That is one of the central civil liberties concerns raised by mobile biometric immigration tools.
The system is not limited at the point of collection to people already known to be removable. It is designed for use before citizenship, status, or identity are resolved. That means U.S. citizens and lawful permanent residents can be scanned, checked, and retained in the same workflow.
The privacy analysis says CBP will retain all photographs, including non-match photographs and photographs of U.S. citizens and lawful permanent residents, as part of TVS holdings consistent with the TVS Privacy Impact Assessment (PIA) and the ATS System of Records Notice (SORN).
It also says the CBP SORN provides coverage for identity verification until ICE takes an enforcement action, at which point the corresponding ICE SORN would apply.
While the document indicates that a PIA and SORN are required, it also says the system is covered by existing privacy documents, including the Traveler Verification Service PIA, the Automated Targeting System PIA, the Enforcement Integrated Database PIA, and the ATS SORN.
ICE’s Privacy Office ultimately marked the Privacy Threshold Analysis “sufficient currently” and said it would provide coverage for TFM for three years unless the collection or use of personally identifiable information changes.
Instead of publishing a standalone, detailed public assessment for a new local-police mobile face-scanning app, ICE appears to be relying on existing privacy paperwork to cover the new operational use by deputized local officers taking face images in the field to support immigration enforcement.
Lawmakers who have been investigating Mobile Fortify say it requires its own separate PIA and SORN.
The TFM plan sits alongside ICE’s broader expansion of field recording and mobile-device use. ICE’s updated body-worn camera directive governs ICE-approved audio, video, and digital recording equipment used during enforcement activity.
But body-worn cameras and facial recognition apps perform different functions. A body camera records what happened during an encounter. TFM would help identify the person encountered, check that person against federal databases, and potentially convert the encounter into an immigration enforcement action.
DHS’s public AI inventory identifies NEC as the vendor supporting the CBP capability that ICE accesses through Mobile Fortify. NEC is one of the world’s major biometric technology companies and has long provided facial recognition and identity systems to government agencies.
DHS’s own disclosures have shown that several key safeguards were still incomplete after Mobile Fortify went operational. The public AI inventory listed a number of risk management steps as still in progress, including an AI impact assessment, independent review, ongoing monitoring, fail-safe procedures, and an appeal process.
Those gaps are especially significant because Mobile Fortify, like TFM, is not limited to noncitizens.
A privacy threshold analysis for Mobile Fortify said users may collect information about people regardless of citizenship or immigration status. It also said people scanned in the field are not given an opportunity to decline or consent.
The analysis said photographs and fingerprints collected through the tool may be retained for 15 years, including data collected from U.S. citizens.
The newly surfaced TFM document shows that same model moving outward to local police.
Civil liberties groups have warned that such systems could turn ordinary police encounters into biometric immigration checkpoints. Their concern is not only that facial recognition can be wrong, but that the consequences of a wrong match in an immigration setting can be severe.
There are already examples of mobile DHS biometrics being used on people who say they are U.S. citizens.
It remains unclear how local and state police officers would be trained to use TFM, how match results would be audited, whether people could challenge an erroneous match, what standards would govern a detention or referral based on an app response, how nonmatch images would be segregated or deleted, and whether local agencies would be required to disclose their use of the tool to the public.
The TFM Privacy Threshold Analysis also raises questions about transparency. The ICE Privacy Office concluded that existing PIA and SORN coverage was sufficient for the time being, even though the system allows local officers to collect biometric and location information from people whose citizenship is unknown.
That leaves the public with only a narrow privacy threshold document for a potentially large expansion of federal-local biometric enforcement.
Members of Congress have already begun pressing DHS on Mobile Fortify. Senators Ed Markey, Ron Wyden, Jeff Merkley and others have asked ICE to stop using the tool and answer questions about testing, accuracy, legal review, retention, and safeguards.
Separate legislation introduced in Congress would restrict or ban ICE and CBP use of facial recognition and other biometric identification systems or sharply limit mobile biometric collection outside ports of entry.
The TFM document is likely to intensify that scrutiny. Mobile Fortify was already controversial because it gave federal agents a way to scan faces and fingerprints in the field. Extending a related capability to local police would move the system into a far larger and less centralized enforcement environment.
The issue now is no longer whether ICE can use mobile facial recognition. It already can. The issue is whether DHS will be allowed to federate that capability outward through local police departments across the country, turning ordinary policing into a front end for federal biometric immigration enforcement.
Article Topics
biometric identification | biometrics | facial recognition | ICE - U.S. Immigration and Customs Enforcement | law enforcement | mobile app | mobile biometrics | Mobile Fortify







Comments