ICE Wants to Know If You’re Posting Negative Things About It Online

Amid anger and protest over the Trump administration’s plan to deport millions of immigrants, U.S. Immigration and Customs Enforcement plans to monitor and locate “negative” social media discussion about the agency and its top officials, according to contract documents reviewed by The Intercept.

Citing an increase in threats to ICE agents and leadership, the agency is soliciting pitches from private companies to monitor threats across the internet — with a special focus on social media. People who simply criticize ICE online could pulled into the dragnet.

“In order to prevent adversaries from successfully targeting ICE Senior leaders, personnel and facilities, ICE requires real-time threat mitigation and monitoring services, vulnerability assessments, and proactive threat monitoring services,” the procurement document reads.

If this scanning uncovers anything the agency deems suspicious, ICE is asking its contractors to drill down into the background of social media users.

That includes:

“Previous social media activity which would indicate any additional threats to ICE; 2). Information which would indicate the individual(s) and/or the organization(s) making threats have a proclivity for violence; and 3). Information indicating a potential for carrying out a threat (such as postings depicting weapons, acts of violence, refences to acts of violence, to include empathy or affiliation with a group which has violent tendencies; references to violent acts; affections with violent acts; eluding [sic] to violent acts.”

It’s unclear how exactly any contractor might sniff out someone’s “proclivity for violence.” The ICE document states only that the contractor will use “social and behavioral sciences” and “psychological profiles” to accomplish its automated threat detection.

Once flagged, the system will further scour a target’s internet history and attempt to reveal their real-world position and offline identity. In addition to compiling personal information — such as the Social Security numbers and addresses of those whose posts are flagged — the contractor will also provide ICE with a “photograph, partial legal name, partial date of birth, possible city, possible work affiliations, possible school or university affiliation, and any identified possible family members or associates.”

The document also requests “Facial Recognition capabilities that could take a photograph of a subject and search the internet to find all relevant information associated with the subject.” The contract contains specific directions for targets found in other countries, implying the program would scan the domestic speech of American citizens.

The posting indicates that ICE isn’t merely looking to detect direct threats of violence, but also online criticism of the agency.

As part of its mission to protect ICE with “proactive threat monitoring,” the winning contractor will not simply flag threatening remarks but “Provide monitoring and analysis of behavioral and social media sentiment (i.e. positive, neutral, and negative).” This includes regular updates on the “total number of negative references to ICE found in social media” from week to week.

“ICE’s attempt to have eyes and ears in as many places as we exist both online and offline should ring an alarm.”

Such sentiment analysis — typically accomplished via machine-learning techniques — could place under law enforcement scrutiny speech that is constitutionally protected. Simply stated, a post that is critical or even hostile to ICE isn’t against the law.

“ICE’s attempts to capture and assign a judgement to people’s ‘sentiment’ throughout the expanse of the internet is beyond concerning,” said Cinthya Rodriguez, an organizer with the immigrant rights group Mijente. “The current administration’s attempt to use this technology falls within the agency’s larger history of mass surveillance, which includes gathering information from personal social media accounts and retaliating against immigrant activists. ICE’s attempt to have eyes and ears in as many places as we exist both online and offline should ring an alarm for all of us.”

Scanning online speech in the name of homeland security is a bipartisan initiative. The document soliciting contractors appears nearly identical to a procurement document published by ICE in 2020, which resulted in a $5.5 million contract between the agency and Barbaricum, a Washington-based defense and intelligence contractor. A new contract has not yet been awarded. ICE spokesperson Jocelyn Biggs told The Intercept, “While ICE anticipates maintaining its threat risk monitoring services, we cannot speculate on a specific timeline for future contract decisions.”

ICE already has extensive social media surveillance capabilities provided by federal contractor Giant Oak, which under both Trump’s first term and the Biden administration sought “derogatory” posts about the United States to inform immigration-related decision-making. The goal of this contract, ostensibly, is focused more narrowly on threats to ICE leadership, agents, facilities, and operations.

Civil liberties advocates told The Intercept the program had grave speech implications under the current administration. “While surveillance programs like this under any administration are a concerning privacy and free speech violation and I would fight to stop them, the rhetoric of the Trump administration makes this practice especially terrifying,” said Calli Schroeder, senior counsel at the Electronic Privacy Information Center. “Threats to ‘punish’ opponents or deport those exercising 1st Amendment rights combine with these invasive practices to create a real ‘thought police’ scenario.”