ICE's New AI Surveillance Tool Reads 8 Billion Social Posts Daily
U.S. Immigration and Customs Enforcement just signed a $5.7M contract for Zignal Labs' AI-powered mass surveillance platform, used by both the Pentagon and Israeli military.
The $5.7 million, five-year contract signed by U.S. Immigration and Customs Enforcement (ICE ↗) with government technology middleman Carahsoft Technology might sound like typical federal procurement bureaucracy. But what’s actually happening is far more sophisticated, and terrifying. Zignal Labs’ platform “analyzes over 8 billion social media posts per day”, quietly ushering in the era of AI-powered mass surveillance.
When Your Social Feed Becomes an Investigation Feed
This isn’t just another government software purchase. The Zignal platform represents a seismic shift in how law enforcement tools scale. While traditional surveillance required human agents monitoring specific targets, Zignal’s AI can process the equivalent of every social media post from every American, twice over, every single day. That’s not monitoring, that’s digital trawling at industrial scale.
The confidential company pamphlet publicly available online ↗ advertises that Zignal Labs “leverages artificial intelligence and machine learning” to provide “curated detection feeds” for its clients, allowing law enforcement to “detect and respond to threats with greater clarity and speed.” But what constitutes a “threat” when you’re processing 8 billion data points daily?
The licenses will be provided to Homeland Security Investigations, ICE’s intelligence unit, to provide “real-time data analysis for criminal investigations”, according to the contract disclosure. But the line between criminal investigation and general population surveillance has never been thinner.
From PR Tool to Pentagon Weapon
Zignal Labs didn’t start as a defense contractor. Founded in Silicon Valley in 2011, the company initially catered to public relations firms and political campaigns ↗, helping clients monitor narrative trends online. That changed dramatically in 2021 when the company formally announced its move into defense and intelligence, complete with a “public sector advisory board” staffed by industry veterans.
The same year, Zignal announced ↗ its strategic pivot toward government contracts. Now, the company’s pamphlet boasts about providing “tactical intelligence” to “operators on the ground” in Gaza for the Israeli military, alongside work with the U.S. Marines and State Department.
This evolution from corporate reputation management to military-grade surveillance reveals a disturbing trend: technologies developed for commercial purposes are increasingly being weaponized for state surveillance operations.
The Surveillance Arsenal Expands
Zignal Labs joins a growing arsenal of digital surveillance tools already deployed by ICE. The agency also uses ShadowDragon, software that maps out an individual’s online activity across publicly available websites, and Babel X, which links social media profiles and location information to a target’s Social Security number.
There’s a pattern emerging: ICE appears to be building what amounts to a social media panopticon. As Wired reported earlier this month ↗, the agency has plans to develop a round-the-clock social media monitoring team to identify leads for immigration enforcers.
This week, ICE signed another $7 million contract ↗ with SOS International LLC for “skip tracing services”, tracking a person’s whereabouts. The timing is notable: just three months after SOS International LLC hired Andre Watson, an ICE Homeland Security Investigations intelligence chief, to “expand [the company’s] business and deliver capabilities to state and federal law enforcement agencies.”
The Human Cost of Automated Enforcement
This isn’t theoretical surveillance, we’re already seeing how these tools translate into real-world operations. Several pro-Palestinian activists, including Mahmoud Khalil, were targeted and jailed by immigration authorities after being doxed online by right-wing, pro-Israel blacklist websites like Canary Mission ↗. Just this week, immigration agents raided street vendors in New York City after a right-wing influencer posted a video ↗ of the block online, demanding action by authorities.
The chilling effect on political speech is precisely what civil liberties advocates warned about. Last week, a group of labor unions sued ↗ over the federal government’s growing use of social media surveillance to target immigrants for their political speech, calling it a “mass, viewpoint-driven surveillance program."
"You Don’t Even Have to Post Personal Information”
The scale of data collection means privacy measures like using pseudonyms provide little protection. As one observer noted, your digital footprint extends far beyond explicit personal information. Your IP address, linguistic style, typing idiosyncrasies, location patterns, online timing, browsing habits, and thousands of other micro-data points create a unique signature that can be tracked regardless of login credentials.
The platform’s AI capabilities mean it can identify patterns and connections that would be impossible for human analysts to detect across such massive datasets. When combined with geolocation data from posts and photos, the system could potentially trace someone’s physical location based on metadata attached to social media content.
The Civil Liberties Battlefield
“[The Department of Homeland Security] should not be buying surveillance tools that scrape our social media posts off the internet and then use AI to scrutinize our online speech”, said Patrick Toomey, deputy director of the ACLU’s National Security Project. “And agencies certainly shouldn’t be deploying this kind of black box technology in secret without any accountability.”
The Electronic Frontier Foundation and Yale Law School’s Media Freedom and Information Access Clinic articulated the core concern in their lawsuit: “The government’s utilization of AI and automated tools for viewpoint-driven online surveillance gives teeth to its threat to surveil ‘everyone’ online for disfavored expression”, they wrote, saying such tools “exacerbate the chilling impact of that surveillance.”
Mission Creep in Digital Dragnets
The most concerning aspect of this technological evolution is what legal experts call “mission creep.” Initially justified for counterterrorism or serious criminal investigations, these surveillance tools inevitably expand to encompass broader applications.
As Julie Mao, an attorney with Just Futures Law, told reporters: “We’ve been seeing an uptick in ICE surveillance contracts.” The pattern is clear, each new capability becomes normalized, then expanded.
The Zignal contract represents more than just another government software purchase. It’s the institutionalization of AI-powered mass surveillance at a scale previously unimaginable. When a single system can process 8 billion posts daily, more than the entire U.S. population could generate, we’ve crossed from targeted monitoring into universal surveillance.
The Zignal contract, made public in September but receiving little attention until now, represents a tipping point in the relationship between citizens, their data, and government power. What was once the domain of science fiction, automated mass surveillance systems parsing our every digital utterance, is now operational federal policy.
The technology’s use by both the Pentagon and Israeli military underscores its battlefield-to-homeland security pipeline. Features developed for identifying threats in conflict zones are now being deployed against domestic populations under the guise of immigration enforcement.
As AI systems grow more sophisticated, the line between public safety and pervasive surveillance becomes increasingly blurred. The $5.7 million question is whether democratic oversight can keep pace with technological capability. Right now, the machines are winning.
