U.S. 🇺🇸 Immigration and Customs Enforcement’s Enforcement and Removal Operations has awarded a contract to a company called AI Solutions 87 for “AI agents” meant to automate skip tracing and network mapping, a move that procurement records suggest will speed up how the agency finds people on its enforcement docket. The contract details were reported in procurement documents reviewed by 404 Media, and they point to a wider turn toward automated immigration enforcement at a time when many immigrant families already live with the daily fear that a traffic stop or a knock at the door could change everything.
What the contract will do
According to 404 Media, the contract is built around feeding identity and location clues into software that can pull from:
- public records
- commercial databases
- social media
- other open-source information
Instead of an officer or analyst working one person at a time, the tools are meant to generate “actionable insights” that support risk-based targeting, including likely places a person lives, works, or spends time with others. The procurement language describes compressing the timeline between identifying someone and carrying out an arrest attempt—taking work that once required long manual searches and pushing it through automated steps.
How the “AI agents” work
The records describe AI agents that can work at scale, connecting names, addresses, phone numbers, and other identifiers into patterns.
- Skip tracing: finding someone’s current location when records are stale or the person does not want to be found — can be done faster.
- Network mapping: linking a person to associates, shared addresses, or other ties — can be done more broadly.
The contract also signals how enforcement increasingly relies on the data trails people leave in ordinary life, from utilities and property records to online profiles and commercial data brokers.
Integration with operations
Procurement material reviewed by 404 Media says automated output will feed into:
- ICE case management platforms
- 24-hour contractor-staffed targeting centers
- Field operations
This matters because it shows analysis being built directly into operations that can run around the clock, not only during normal government office hours. It also suggests a growing role for private contractors in core enforcement tasks, with software and staff sitting between raw data and officers in the field.
The same records describe cross-referencing addresses and quickly identifying associates, which can widen the circle of people whose information gets pulled into a case file.
Related systems and partnerships
This contract sits alongside other systems and partnerships that expand ICE’s data and surveillance capabilities:
| System / Partner | Purpose / Capability |
|---|---|
| ImmigrationOS | Provides near real-time visibility into a person’s status and movements across the immigration lifecycle (prototype referenced in source material). |
| Palantir Technologies | Partnership tied to ImmigrationOS; tools used to track immigrants, analyze driver’s license data, extract info from seized phones, and cross-search location data. |
| U.S. Border Patrol — CMPRS | Conveyance Monitoring and Predictive Recognition System scans license plates nationwide and flags travel patterns deemed suspicious. |
| Flock Safety | Automated license plate reader (ALPR) data reportedly shared by local agencies with Border Patrol. |
The stated goal across these tools is speed and scale: more data sources, faster linking, and quicker output pushed toward enforcement teams.
Effects on immigrant communities
For immigrants and their families, the shift from human-led research to automated searches can feel less like an efficiency upgrade and more like an expansion of the state’s ability to find them wherever they go.
- When location is inferred from many sources at once—work histories, address histories, online posts, and commercial records—people who move often, live in shared housing, or work informal jobs may worry that a friend’s address or a relative’s phone number could pull them into the same network map.
- The procurement documents do not include direct quotes from ICE officials or the contractor, but the contract descriptions emphasize producing leads that are ready for action.
Key takeaway: Automated linking of everyday data points increases the likelihood that routine associations (shared housing, a friend’s phone number) will be used as enforcement leads.
Similar automation trends in border policing
The source material shows the automation trend is not limited to ICE:
- U.S. Border Patrol’s CMPRS scans license plates nationwide and uses algorithms to flag travel patterns based on origins, destinations, and routes, which can lead to stops, searches, and arrests.
- Customs and Border Protection requested $2.7 billion in FY2024 for AI-enhanced border surveillance.
- Wide-area license plate scanning and pattern analysis can reach deep into the interior, raising questions for communities far from the border about when routine travel becomes a trackable data stream.
Local data sharing and access methods
The source material reports specific examples of local-to-federal data sharing:
- In Washington state, at least eight local agencies shared Flock Safety ALPR data directly with Border Patrol in 2025.
- Ten other agencies reportedly allowed “back door” access from May to August 2025, enabling immigration-related searches despite state protections.
The language referencing “front,” “back,” and “side door” methods highlights that access to surveillance data can move through many channels, including routine agreements local agencies may treat as normal information sharing. It also underscores how tools bought for local policing can be repurposed by federal enforcement.
Risks, errors, and civil rights concerns
Civil rights advocates warn that automation can make mistakes harder to catch, especially when large datasets contain:
- outdated addresses
- shared names
- incorrect links between people
Even without new laws, higher-speed targeting can change on-the-ground reality:
- A wrong address can cause a frightening encounter for someone uninvolved.
- A correct address can turn an ordinary routine—school drop-off, a shift change, grocery run—into the moment enforcement arrives.
VisaVerge.com reports that automated targeting raises practical problems for immigrants trying to keep their lives stable, because a single data point—an old lease, a past utility bill, or an online account—can follow someone for years.
Historical context and continuity
The source material frames this as part of a longer arc. ICE tactics such as data surveillance, local jail information sharing, and ruses have been used for years.
- 2023 surveys by the Immigrant Legal Resource Center found these tactics persisted under Biden priorities.
- The reports note no major changes into 2025.
This continuity matters: technology changes, but the underlying goal—finding and arresting people on an enforcement docket—remains steady even as administrations describe priorities differently.
Practical guidance for families trying to locate detained loved ones
For families trying to locate someone already in custody, the most reliable channel remains ICE’s official locator.
- ICE advises the public to use the Online Detainee Locator System, which allows searches by a person’s name and birth details or by an A-number.
- ICE also directs people to contact local ERO field offices.
Official tool: ICE Online Detainee Locator System
As contracts like the AI Solutions 87 award show how much enforcement is being shaped by data systems, detainees and their relatives are often left to rely on a single official portal to find out where a loved one is being held and what happens next.
ICE has contracted AI Solutions 87 to implement “AI agents” designed to automate the process of finding and tracking immigrants. By integrating public, commercial, and social media data, these tools speed up “skip tracing” and “network mapping.” While the agency seeks operational efficiency and faster arrests, the shift raises significant concerns regarding privacy, data accuracy, and the impact of automated surveillance on immigrant families.
