Bipartisan consensus is rare these days. The phrase itself evokes a bygone era like dial-up internet. That said, immigration policy was one of the few issues where Vice President Harris and President-elect Trump found some common ground this past cycle.
Harris moved to the right to appear tough on immigration, losing to Donald Trump who ran in part on the promise of widespread deportation. It’s not a partisan issue; elected officials on both sides of the aisle support a brutal immigration policy.
Beyond just the brutality, there is something to be gleaned about the future of the United States by looking at the opaque public-private partnerships that make up the immigration surveillance state.
A network of personal information databases, biometric data and potent spyware make up the immigration surveillance octopus, with each tentacle representing a broad and distinct strategy to combat what is seen as an existential threat to the ole’ Red White and Blue.
A look at three arms of the American immigration surveillance network give a better understanding of what the future holds: the VLVI program (formerly called extreme vetting) used by the Department of Homeland Security to surveil immigrants’ digital footprints for visa and entry status, Clearview AI’s biometric surveillance for immigration officials. and Israeli spyware maker Paragon’s partnership with Immigration and Customs Enforcement.
The artist formerly known as extreme vetting
In the earliest days of the first Trump administration, before the backstabbing tell-alls and cabinet swaps, then-President Trump issued a bombastic executive order that banned travel to the US for 90 days from seven predominantly Muslim countries in what quickly became known as the Muslim Ban.
Language used in that order then became the criteria for the sweeping DHS surveillance program, costing U.S. taxpayers north of $100 million and restricting first amendment rights, now called the Visa Lifecycle Vetting Initiative.
Both ICE and the DHS use VLVI to track migrants’ online presence across a variety of social media sites, ostensibly to determine if said immigrants are a threat, leading to potential deportation or visa denial. The purview of the program as stated in documents I have reviewed
was broad, with no clear definition of what constitutes a threat.
In August 2018, DHS contracted SRA International (a subsidiary of General Dynamics that has since changed its name to CSRA Inc.) to:
“…analyze and apply techniques to exploit publicly available information, such as media blogs, public hearings, conferences, academic websites…”
The bulk of this information comes from the Electronic Frontier Foundation and Aaron Mackey, the free speech and litigation director there, who filed a FOIA lawsuit against the DHS in 2022 for more information on the then-called extreme vetting program.
According to EFF’s reporting, DHS entered into a contract with SRA International from the early days of the Trump Administration, August 2018, through the waning days of the Biden Administration in August 2023 for $42.1 million. The complaint also details a $4.8 million payment to SRA on May 4, 2021.
That appears to be just the first part of the deal. Based on the documents from the EFF complaint, the SRA International contract ultimately cost the taxpayers $101 million.
When I spoke to Mackey he offered a revealing, if not frightening, picture of the program and highlighted how little is publicly known about it even now.
“…so there’s some sort of automated system or tool that’s sort of an open source ingesting of social media feeds and publicly available websites. And also, I think it collects press reports. Think of it like a giant firehose and identifies everything, breaks it apart and then allows for keyword searches and allows it to be queried,” he said.
One of the many pitfalls of this type of automation that Mackey touched on is that these systems have proven to be prone to errors which are hard to correct.
“Automated systems are really poor at basic human functions like reading context humor… On top of that, you’ve got this huge error issue where the government has this broad discretion to decide what is and what isn’t derogatory information,” Mackey said.
Beyond just that, the individual being targeted not only has no recourse to challenge the decision but is often unaware of the role this automated system plays in a monumental decision like visa status. Mackey called on elected officials to take a more active role in protecting people’s First Amendment rights.
“You may never learn that it [the system] made this decision. What we need is the folks in these agencies and congress and the executive branch to really come to grips with that and take seriously their duties to protect not only people’s first amendment rights but their rights to have this process play out in a fair way,” he said.
Mackey highlighted existing examples within the national security state to illuminate both the opacity of the technology and the difficulty individuals face when trying to remedy the error.
“Once a mistake and error is included in the system, they’re generally hard to undo. And it’s possible, even highly likely that the person who’s had adverse action taken against them is not likely to know that it’s because of an error,” he told me.
Mackey used the example of the no-fly list that often mistakenly targets people with the same name as someone the list. As he explains, it’s difficult to fix the error once someone is caught in the dragnet because “the error permeates a whole bunch of other government systems and officials.”
The errors in these systems are one issue. Another is the ambiguity of the purview, as it gives SRA and any other contractor with the VLVI program powers to surveil social media profiles for “derogatory information” (the term used in the bid) without any time limitations.
“I think the idea is, it’s sort of an ongoing process where there’s some ability to surveil and monitor their online activities…So, it’s not a process of being screened once and then everything is okay, it’s this ongoing surveillance is what it looks like,” Mackey said.
Setting all of this aside just for a moment, the premise of this kind of program butts up against what we’re taught is a First Amendment right: freedom of speech.
“You’re looking at a system that is primed for errors, misuse, and the ambiguity could lead to a lot of decisions that are made based on people exercising what we’d think of as First Amendment activity. I think that’s incredibly problematic when the scale of this system is so broad and it goes on for the whole time that someone is a visa applicant or visa holder,” Mackey told me.
Clearview’s all-seeing eye
Another tentacle in the immigration octopus is biometric data. For the laypeople like me, that’s basically just your faceprint. Clearview AI, a startup founded by Australian-Vietnamese national Hoan Ton-That and New York political operative Richard Schwartz, has made a killing selling your faceprint.
The company made a splash during the early Trump admin over alleged privacy violations and ties to right-wing pundits but has since caught a massive wave of government contracts.
In Feb. 2020, Buzzfeed reported that more than 2,000 law enforcement agencies, private companies and government wealth funds were clients of Clearview AI. Later that year, reporting originally from Tech Inquiry republished by The Verge revealed that ICE signed a contract with Clearview for $224,000 to license the technology with the ICE Dallas as the contracting office.
No success comes without haters though and Ton-That, Schwartz and Clearview have been embroiled in multiple lawsuits including a class action in Illinois, as well as a FOIA lawsuit filed by Just Futures, Immigrant Defense Project, Mijente and the ALCU of Northern California in April. 2021.
That joint lawsuit hinges on the contracts that Clearview has with ICE and the DHS. Notably, some California law enforcement departments are named in the suit as they have previously contracted or are currently contracting Clearview for the biometric surveillance. The California jurisdictions include the County of Alameda, the City of Alameda, the City of El Segundo, San Luis Obispo County Sheriff’s Office and the Lake County Sheriff’s Office.
Considering Ton-That marketed the product as a tool for law enforcement, it’s not surprising to see CA departments jumping on board. It is, however, toeing the line of illegality as Assembly Bill 1814, passed in July 2024, prohibits the use of facial recognition technology as the sole basis for probable cause in a criminal complaint. The phrasing allows for FRT to be used in conjunction with other evidence.
In September of this year, Just Futures issued an objection to the Clearview class action settlemen, based on two components.
“We objected both on the grounds that the settlement agreement doesn’t address the harm that’s ongoing for people but also the fact that people don’t know about this and we don’t feel that the notice plan that’s included in the agreement goes far enough to make sure that people really understand the stakes of what they’re giving up potentially,” Dinesh McCoy, the Just Futures staff attorney who authored the objection told me in an interview.
While ICE initially claimed that Clearview would exclusively be used for child exploitation or cybercrime, the software’s purview was broadly expanded to include identity theft crimes, according to documents released in conjunction with the Mijente, Just Futures lawsuits. Additionally, those documents also detail the National Leader Development Center’s use of Clearview for potential fraud cases regarding immigrant benefits.
McCoy noted the dangers of partnerships that give private companies broad surveillance powers.
“Private companies don’t have the same legal restrictions that the government has in terms of collecting information and so private companies don’t need a warrant to gather all sorts of information about you. They don’t have a Fourth Amendment restriction … so they’re out there getting this information wherever they can and that’s the business model now,” he said.
McCoy echoed Mackey’s point that beyond the broad powers these companies now have, the actual collection and use of the information collected is opaque, with individuals caught in the dragnet rarely, if ever, getting a chance to contest the ruling made by an automated system that changes the course of their lives.
“ …really pinpointing how those technologies are being used into any specific person’s life is really difficult. There’s not often a requirement in ICE policy or the investigative process that will make it so that you’ll know if … this technology was used or Clearview was used,” McCoy told me.
Paragon’s partnership
Law enforcement’s third tentacle of surveillance is spyware. This is one of the most active parts of the octopus as spyware is meant not just for collection from available or semi-public data but rather for extracting information from more clandestine sources.
Despite then President Joe Biden’s executive order prohibiting the use of commercial spyware, in early October, 2024, ICE signed a $2 million contract with Paragon Solutions, the American subsidiary of the Israeli spyware maker, according to Wired.
That was not Paragon’s first government contract. The New York Times reported that the DEA began using the Graphite program, designed primarily for extracting data from cloud backups, in 2022. In 2021, Forbes reported that the Graphite program was used to decrypt communications on encrypted apps like Signal and WhatsApp. It’s not clear if the ICE contract was for Graphite as well.
In December of 2024, multiple outlets reported that Paragon was sold to American private equity firm AE Industrial Partners. Reports differ on the amount though the range appears to be anywhere from $450 million to $900 million, based on the company’s growth.
Paragon was founded by former Israeli intelligence official Ehud Schneorson and former Israeli Prime Minister Ehud Barack, Tech Crunch and YNet News have reported. Schneorson is a retired brigadier general and the former commander of Unit 8200, which some publications have likened to the US National Security Agency.
Reuters reported the unit was involved in the infamous pager attack in Lebanon that injured hundreds and killed at least nine people, including children. The New York Times reported that Unit 8200 was using Google Photos and Foresight in a facial recognition program to surveil Palestinians in March 2024. Beyond that, the Wall Street Journal reported in August 2024 that Unit 8200 has become Silicon Valley’s “talent pipeline.”
Despite the contrarian points that this is all well established, privacy is a thing of the past like dial-up, and that folks who aren’t doing anything wrong, don’t have anything to hide, there’s a broader takeaway here.
What is to be done?
Bipartisan consensus does still exist in our country but only for war. America’s wide reaching surveillance octopus is used for immigrants, but it is just as often turned against American Blacks, Muslims, or political radicals. While Democrats and Republicans squabble over money for education or Medicaid, there’s always money for more weapons, spyware or surveillance.
Beyond the bipartisan consensus, the way in which the money is distributed is instructive. As shown by the veritable Russian nesting doll of contractors detailed above, a huge swathe of the government’s leg work regarding immigration is handled by private companies. These companies’ broad data collection and extraction powers don’t stop at immigrants; the tentacles reach into everyone’s lives. Without creating some kind of guard rails for what is and isn’t allowed in terms of collection, we’re leaving our future to be determined by Clearview, Paragon, SRA or the lowest bidder.
Thus, the work of EFF, Just Futures, Mijente and similar organizations are a critical bulwark against government and corporate overreach. Left unabated, these companies we’ll scrap every part of the daily human experience for profit. We can styme their powers, but only by working collectively with these organizations fighting for our dwindling privacy rights.
