ICE’s contracting trail shows the rise of automated immigration enforcement

The revelation by 404 Media that Immigration and Customs Enforcement (ICE) quietly contracted with a company advertising “bounty hunter AI agents” did not expose an outlier experiment so much as illuminate a specific capability embedded within a broader enforcement architecture ICE has been assembling through procurement.

According to contract records reviewed by 404 Media, ICE’s Enforcement and Removal Operations awarded a contract to AI Solutions 87 for services marketed as “AI agents” designed to assist with locating people on ICE’s enforcement docket.

While modest in dollar value when viewed in isolation, procurement filings indicate that the contract’s function – automated skip tracing and network mapping – aligns closely with other ICE efforts to expand data-driven targeting and location capabilities.

Procurement records and the company’s own marketing materials describe AI Solutions 87’s software as capable of rapidly identifying persons of interest, mapping their associates, and surfacing likely locations. ICE has not publicly disclosed technical specifications for the system, including how the software operates internally or what data sources it relies upon.

Based on the scope described in the contract and vendor materials, however, what ICE appears to be purchasing is not an autonomous system operating independently of human oversight, but a specialized analytical tool designed to accelerate identity resolution and skip tracing, tasks that have traditionally required substantial manual effort by investigators and private contractors.

That distinction matters, but it does not diminish the significance of the contract. When considered alongside ICE’s broader procurement activity over the past year, the AI Solutions 87 agreement appears less like a pilot and more like the insertion of an automated “find” function into an already industrialized enforcement pipeline.

Procurement filings and agency justifications show ICE systematically investing in systems that shorten the distance between identifying a person in a database and locating that person in the physical world.

Throughout 2025, ICE documents describe a sustained effort to build persistent, large-scale target discovery capabilities. Requests for information and Statements of Objectives circulated by the agency for “Data Subscription and Analytical Services” indicate ICE is seeking enterprise-wide platforms capable of continuous monitoring and alerting across vast populations of people or entities of interest.

These procurement documents describe services that would fuse public records, proprietary commercial databases, social media, and other open source information into analytical environments designed to produce “actionable insights” and support risk-based enforcement prioritization.

While framed in bureaucratic terms, the filings indicate ICE is pursuing systems that actively watch, correlate, and flag targets at scale, rather than merely storing records.

At the center of this ecosystem sits ICE’s investigative case management infrastructure, most notably the Palantir-built Investigative Case Management system and its evolving derivatives.

ICE documents describe these platforms not simply as repositories of information but as workflow engines. Limited sources justifications and related reporting indicate the systems are designed to link identities across databases, deconflict subjects between cases, support structured and unstructured searches, and route leads through internal review and enforcement pipelines.

Procurement filings describing the ImmigrationOS prototype explicitly frame new capabilities as providing near real-time visibility into individuals’ status or movements and as supporting targeting and enforcement prioritization across the immigration lifecycle.

From this documented foundation, the AI Solutions 87 contract appears to add speed and automation at the front end of the process. Skip tracing – determining where someone is likely to be living, working, or associating – has historically depended on time-intensive research, database queries, and manual verification.

The contract and vendor descriptions reviewed by 404 Media indicate that AI Solutions 87’s software is intended to automate portions of that work by ingesting identity data and rapidly mapping networks and potential locations, thereby compressing the timeline between identification and enforcement action.

That function becomes more consequential when viewed alongside ICE’s expanding reliance on contractor labor. Reporting over the past year shows ICE planning and executing contracts for private analysts to staff 24-hour targeting centers, where social media and other open source intelligence are converted into enforcement leads.

Procurement materials and reporting describe contractors using commercial identity-resolution tools to assemble dossiers that are then routed to field offices.

Other ICE contracts indicate a parallel expansion of field-level verification and surveillance services, with private firms tasked with confirming addresses, photographing residences, and documenting daily routines at scale.

Within this pipeline, AI Solutions 87’s “agents” occupy a narrow but high-leverage role. They do not make arrest decisions or conduct physical surveillance. Instead, they generate and enrich leads that drive downstream activity.

A system that can more quickly identify associates, cross-reference addresses, and flag likely locations increases the throughput of the entire enforcement apparatus. Procurement filings and related reporting suggest these outputs are designed to feed directly into case management platforms, contractor-run targeting centers, and ultimately into field operations.

This is why the “bounty hunter AI” framing, while rhetorically charged, is not entirely divorced from the operational reality described in ICE’s contracting record. ICE has already outsourced key enforcement-adjacent functions LIKE analysis, surveillance, and physical location work to private firms.

What procurement filings indicate is new is the degree to which automation is now being layered onto that outsourcing, reducing friction and expanding capacity. The “bounty hunter” aspect lies less in the software itself than in how its outputs are operationalized by contractors paid to locate people efficiently and repeatedly.

From an oversight and civil liberties perspective, the convergence raises familiar but intensifying concerns. Automated skip tracing relies heavily on commercial and open source data that can be inaccurate, outdated, or misleading.

Network mapping tools can pull in family members, roommates, or casual associates who have done nothing wrong, yet find themselves part of an enforcement dossier.

Because these processes are driven by proprietary vendor systems, ICE documents provide little public insight into how targets are identified, prioritized, or challenged.

Procurement justifications also reveal a pattern of institutional dependency. ICE filings repeatedly argue that certain vendors are uniquely positioned to deliver new capabilities because their systems are already embedded as systems of record.

Once accepted, that logic can turn limited or prototype deployments into permanent infrastructure, not through public debate or legislative authorization, but because replacing them would slow operations.

Seen in this light, the AI Solutions 87 contract does not represent a radical departure for ICE, but rather it reflects a continuation of a strategy documented across multiple procurements in which ICE is replacing labor-intensive investigative work with data-driven automation, outsourcing both analysis and action to private firms, and stitching the system together through centralized case-management platforms that translate information into enforcement outcomes.

Article Topics

AI agents  |  automation  |  border security  |  CBP  |  ICE - U.S. Immigration and Customs Enforcement  |  immigration  |  ImmigrationOS  |  procurement  |  U.S. Government