Amazon aggressively pitches Rekognition facial recognition to ICE during family separation crisis

| Importance: 9/10 | Status: confirmed

In June 2018, at the height of the Trump administration’s family separation crisis at the U.S.-Mexico border, Amazon Web Services officials met with Immigration and Customs Enforcement (ICE) representatives in Redwood City, California to pitch Rekognition facial recognition technology for immigration enforcement operations. The meeting, held at consulting giant McKinsey & Co.’s offices and revealed months later through Freedom of Information Act requests by the Project on Government Oversight (POGO), demonstrated Amazon’s aggressive pursuit of immigration enforcement contracts despite mounting criticism of both ICE’s practices and facial recognition technology’s documented racial bias.

The Sales Pitch

An Amazon “sales principal” sent a follow-up email to ICE’s Homeland Security Investigations (HSI) office on June 15, 2018, describing the meeting and outlining “action items” including setting up a technical briefing for ICE officials about Rekognition’s capabilities. The email, later published by The Daily Beast, stated: “We are ready and willing to support the vital Homeland Security Investigations mission.”

The pitch emphasized Rekognition’s tagging and analysis capabilities, including its real-time facial matching system. Amazon officials told ICE that the software could be used during immigration investigations to identify and track undocumented immigrants. The cloud-based system would enable ICE to process vast amounts of surveillance imagery without investing in expensive local infrastructure.

Timing and Context: Family Separation Crisis

The timing of Amazon’s sales efforts proved particularly controversial. The June meeting occurred during the Trump administration’s “zero tolerance” immigration policy that resulted in the forced separation of thousands of migrant children from their parents at the southern border. Images of children in cages and audio of crying children separated from parents had sparked national outrage and international condemnation.

Amazon pitched its facial recognition software to the Department of Homeland Security “at the height of the family separation crisis,” as Common Dreams reported, offering the system as one that could help agents identify and detain undocumented immigrants—precisely the operations that were tearing families apart.

Employee Protest and Internal Opposition

Roughly one week after the ICE meeting, hundreds of anonymous Amazon workers wrote a letter to CEO Jeff Bezos stating: “We refuse to build the platform that powers ICE, and we refuse to contribute to tools that violate human rights.” The letter specifically mentioned outrage over ICE’s separation of migrant parents and children at the Mexican border.

The employee letter represented a significant moment of internal dissent, with Amazon workers explicitly connecting their labor to human rights violations. Despite these protests, Amazon continued pursuing the ICE contract and defending Rekognition’s use by law enforcement and immigration agencies.

ACLU and Civil Liberties Response

The revelations about Amazon’s ICE pitch intensified criticism from civil liberties organizations. The ACLU called for a moratorium on law enforcement and immigration enforcement use of facial recognition technology, citing concerns about accuracy, racial bias, and the potential for mass surveillance of immigrant communities.

The ACLU noted that Amazon was simultaneously marketing Rekognition to local police departments across the country, creating an interconnected surveillance infrastructure that could be used to identify and track immigrants in their communities, workplaces, and homes. This raised concerns about the creation of a nationwide dragnet that could enable mass deportations and chill First Amendment activities in immigrant communities.

Broader AWS-ICE Relationship

The Rekognition pitch was part of Amazon’s broader relationship with immigration enforcement agencies. As MIT Technology Review documented, Amazon Web Services provided the “invisible backbone” of ICE’s immigration crackdown, hosting critical databases and providing cloud computing infrastructure that powered ICE’s operations. The addition of facial recognition capabilities would have significantly expanded ICE’s surveillance and enforcement capacity.

While there is no evidence that ICE ultimately purchased Rekognition—an ICE spokesperson stated that publicly available procurement data showed no contract with Amazon for facial recognition services—the aggressive sales pitch demonstrated Amazon’s willingness to provide surveillance technology to one of the most controversial federal agencies at the height of its most criticized operations.

Significance for Surveillance State

Amazon’s ICE pitch represented a troubling convergence of corporate surveillance technology and immigration enforcement during a period of documented human rights abuses. The episode illustrated how tech companies were actively marketing AI-powered surveillance tools to agencies conducting controversial enforcement operations, prioritizing revenue opportunities over ethical concerns about how the technology would be deployed.

The revelation that Amazon pursued this contract despite employee protests and during the family separation crisis highlighted the company’s commitment to expanding its government surveillance business regardless of civil liberties implications. It also demonstrated how facial recognition technology was being marketed not as a narrow tool for specific law enforcement purposes, but as a general surveillance capability for identifying and tracking entire populations—in this case, immigrant communities already subject to intense scrutiny and enforcement pressure.

The episode became a key example in debates about tech industry complicity in government human rights violations and helped fuel growing opposition to facial recognition technology from civil liberties organizations, researchers, and even some lawmakers who began calling for restrictions on the technology’s use by government agencies.

Help Improve This Timeline

Found an error or have additional information? You can help improve this event.

✏️ Edit This Event ➕ Suggest New Event

Edit: Opens GitHub editor to submit corrections or improvements via pull request.
Suggest: Opens a GitHub issue to propose a new event for the timeline.