As artificial intelligence, license plate readers, data brokers and federal immigration databases become increasingly connected, immigrant rights advocates and privacy experts are warning that surveillance technology is rapidly reshaping immigration enforcement across the country.

At a special media briefing on September 10, hosted by American Community Media, a panel of experts described how federal immigration enforcement is being powered not only by agents on the ground, but by vast systems of data collection that can track movement, identify patterns and make it easier for agencies to locate people.

Speakers

  • Denise Bell – Human rights researcher, report co-author, Agents of Chaos and Cruelty
  • Lena Graber – Senior Managing Attorney, Immigrant Legal Resource Center
  • Brian Hofer – Executive Director, Secure Justice
  • Steven Hubbard – Senior Data Scientist, American Immigration Council
  • Saira Hussain – Senior Staff Attorney, Electronic Frontier Foundation
  • Stephen Manning – Executive Director, Innovation Law Lab
  • Jonathan Paz – ex Flock whistleblower, Candidate for Congress from MA 5th District
  • Abdi Soltani – Executive Director, ACLU of Northern California
  • Kimberly Woo – Community Organizer, SIREN Immigrant Rights

(Top, L to R; Bottom L to R)


The central warning was clear: surveillance is no longer limited to a single camera, database or agency. Instead, speakers described an expanding ecosystem in which federal, state, local and commercial data streams can be combined to support immigration enforcement.

American Community Media Executive Director Jaya Padmanabhan opened the convening by noting that the speed at which surveillance tools are being developed has outpaced public understanding and legal oversight.

“The rapid advancement of AI and its deployment in the surveillance industry without meaningful guardrails is genuinely frightening,” she said. “Who is being targeted? When? How? And where, what happens to that data? And how is it used?”

AI and Immigration Enforcement

Steven Hubbard said reporters and the public should not think of AI only as tools like ChatGPT. In government surveillance, he explained, AI is often used for pattern recognition, risk scoring, facial recognition, license plate scanning, language processing and large-scale database searches.

“When you’re reporting about government technology, I wouldn’t focus too much on whether it’s AI or not,” Hubbard said. “A better question is asking, what is the system doing? Right? And what is powering that system?”

Hubbard said the greatest power often comes not from the sophistication of a single AI tool, but from the ability to connect large amounts of data across systems.

“Surveillance power doesn’t necessarily come from a particularly sophisticated AI system,” he said. “It comes from connecting data.”

Immigration agencies, he said, can access immigration records, border crossing records, biometric information, student records and enforcement records. They may also obtain information from other federal agencies, state and local governments and commercial data brokers.

Once those systems are connected, Hubbard said, technology can search for relationships, patterns and associations at a scale that would be impossible manually.

But he warned that those associations can be wrong or misleading.

“A pattern is not necessarily proof. A match is not necessarily identity. And a prediction is not necessarily fact,” Hubbard said.

Operation Black Rose and the Human Impact

Stephen Manning described how surveillance systems can translate into real-world enforcement.

He discussed a case involving Operation Black Rose, a DHS operation in Oregon that he said involved facial recognition and other surveillance tools. Manning showed body camera footage from an early morning enforcement action in Woodburn, Oregon, a largely Latino and immigrant community south of Portland.

At the time agents surrounded a van, Manning said, they did not know who was inside.

“They know exactly no one’s identity in the van,” he said. “They don’t know who was in this van.”

Manning said agents used tools including Palantir’s Elite system, Mobile Fortify, Mobile Query and other data systems to identify and target the vehicle. He described Elite as being compared in testimony to “Google Maps, but for people,” with pins representing individuals.

The case began as a habeas petition and later became a class action. Manning said the lawsuit recently won a motion to compel, allowing broader discovery into DHS surveillance systems and how they contributed to the enforcement action.

For Manning, the case illustrated how digital systems can create what he called a dragnet.

“They created this, this dragnet, using surveillance so that the agent showed up one morning and said, okay, let’s go to this area and see who we can pick up and see who we can find,” he said.

Human Rights Findings

Denise Bell said her report reviewed more than 1,200 immigration enforcement incidents in 2025 across several states, including California, Arizona, Colorado, Florida, Illinois, Louisiana, Maryland and New Mexico.

The research, she said, was based on publicly available information and did not filter specifically for civil rights violations. Instead, researchers asked what happened, who was targeted, where enforcement occurred and what systems made those incidents predictable and repeatable.

Bell said the report found a pattern of civil rights violations across states and communities.

“What we found was a pattern and practice of civil rights violations in state after state in both cities with massive enforcement surges, and in communities that never saw significant federal law enforcement deployment,” Bell said.

The report documented immigration enforcement in everyday places such as stores, streets, sidewalks, parking lots, bus stops, malls, gas stations, parks and wildlife reserves. Bell also said agents used force, threats of force, intimidation and deception as common tactics.

She said the report documented 624 enforcement actions in everyday places, 252 incidents at workplaces such as Home Depots, car washes and restaurants, 432 incidents of misconduct and 437 incidents of likely racial profiling.

Bell said surveillance technology and racial profiling often operated together.

“People were stopped and detained because of perceived race, ethnicity, spoken language, accent, work, or presence in a particular place,” she said.

Data Collection and Digital Rights

Saira Hussain said the federal government is working to consolidate as much information as possible about individuals in order to support immigration enforcement.

That includes information from other federal agencies, state-administered benefit programs, driver databases, voter rolls and commercial data brokers.

Hussain said the government often obtains information through what privacy advocates call the “data broker loophole.” Instead of seeking a warrant, agencies may buy data from private companies that collect and package personal information.

“That is the data broker loophole that we often talk about,” Hussain said.

She warned that when multiple data streams are combined, the result can be a deeply revealing picture of a person’s life.

“This is everything the government has on you,” she said. “It could be very, very revealing.”

Hussain also pointed to social media surveillance of visa holders and protesters as another growing concern. She said some visa applicants are now required to make social media profiles public, while a lack of social media presence can also raise suspicion.

The result, she said, is that people may delete posts, avoid online speech, withdraw from associations or self-censor.

Local Surveillance and Flock Cameras

Jonathan Paz described his time working on the company’s government affairs team from 2024 to mid-2025.

Paz said he initially believed the technology could help recover stolen vehicles and find missing people. But he said he later concluded that automated license plate reader systems and related surveillance tools were being sold in ways that bypassed meaningful public scrutiny.

“Their main role is to surveil the public,” Paz said of companies like Flock.

He described the sales pitch around automated license plate readers as “flockaganda,” arguing that emotionally compelling case studies about missing children or stolen vehicles were used to persuade city councils even when broader questions about surveillance, privacy and immigration enforcement remained unresolved.

Paz said local safeguards and police department policies often give the public a false sense of protection.

“These cameras are gonna go up and once they go up, it’s very hard to control how the local law enforcement agencies use them,” he said. “Because who’s policing the police?”

Paz also said communities should assume that federal agencies may be able to access automated license plate reader networks, especially in areas with both immigration enforcement activity and dense camera coverage.

Brian Hofer said the concern is not just one company or one camera system. The larger problem, he said, is the creation of national surveillance ecosystems.

“It’s never just a license plate reader,” Hofer said. “These have become giant national ecosystems that really can track you throughout the country.”

Hofer said automated license plate readers can capture more than plate numbers. They can also capture vehicle characteristics, travel patterns, associations and other details that allow systems to infer where people go and who they may be connected to.

“That’s never been true, ever,” he said of the claim that the technology is “just a license plate.”

Hofer also said public opposition is having an impact, with many jurisdictions reconsidering or ending Flock contracts. Still, he warned that replacing one vendor with another does not end mass surveillance.

“Unless there is the political will to just stop the practice of mass surveillance altogether, they’re just gonna swap parties,” he said.

Community Organizing in the Bay Area

Kimberly Woo said immigrant communities are already feeling the chilling effect of surveillance.

She said families are afraid that information they shared to access services, benefits or legal processes could later be used against them.

“Federal government has abused our private data to detain and deport our own,” Woo said.

Woo also shared a personal story about loved ones who had waited more than 20 years for permanent residency, only to be detained after reaching the final green card interview. One was deported, she said, while the other later chose to self-deport because of detention conditions.

For Woo, the fear created by surveillance extends beyond those directly targeted.

“Mass surveillance does not keep us safe,” she said. “In fact, it keeps us afraid.”

SIREN joined a lawsuit against San Jose over police access to automated license plate reader data from more than 500 cameras. Woo said community organizing against Flock and automated license plate readers has since expanded across the Bay Area.

In Mountain View, she said, a broad coalition helped secure a unanimous vote to end Flock use and remove the cameras.

“We secured a unanimous victory, to end Flock and remove the cameras,” Woo said.

She said similar organizing has spread across Santa Clara County, the Peninsula, San Francisco, Alameda County, Contra Costa County, Solano County and other Bay Area communities.

Sanctuary Policies and Their Limits

Lena Graber said sanctuary policies still matter, but many were not written to address the newer surveillance systems now being used in immigration enforcement.

She said more than half of ICE arrests nationally happen because someone is transferred from another law enforcement agency. Sanctuary policies that limit cooperation between local police and ICE can reduce those outcomes.

“Sanctuary policies do work,” Graber said.

But she warned that surveillance tools add new risks by helping agencies identify who may be deportable and where they may be found.

“They want to know where I am or where I’m likely to be, so they can arrest me,” she said.

Graber said some forms of immigration enforcement remain old-fashioned: arrests at immigration court hearings, ICE check-ins, airports or outside workplaces. But data systems and surveillance tools are expanding the government’s ability to identify and locate people.

“In a daily level, surveillance is everywhere, and most of it is really beyond the control of an individual person,” she said. “The landscape is constantly changing. So the response that we make really has to be collective.”

A Broader Privacy Fight

Abdi Soltani said California’s state constitution may provide important protections even as federal Fourth Amendment law has weakened.

He pointed to California’s express right to privacy and state protections against unreasonable search and seizure as potential legal tools.

“There have been decades when states led the progress for civil rights, and this is a time when we need California to lead again on civil rights and constitutional issues,” Soltani said.

Soltani said the issue affects everyone because surveillance systems threaten basic freedom of movement. But he said immigrants are especially vulnerable when data is used to support detention and deportation.

“Every person deserves the freedom of movement, the right to move about their community without the government tracking every movement across cities and within cities,” he said.

For speakers at the convening, the fight over surveillance is no longer only about privacy. It is about immigration enforcement, civil liberties, public accountability and whether communities can move through daily life without being constantly tracked.

As Soltani put it, journalism has a key role in exposing systems that are often designed to operate out of public view.

“Journalism is a key strategy through which these types of things that are meant to be secret, can be exposed and made open to the public and open to our democratic oversight.”

 

All images courtesy of American Community Media