How AI is aiding Trump's immigration crackdown

• More law enforcement has access to personal data
• Cases of immigration arrests have been based on inaccurate information
• Local law enforcement deputized to act as federal immigration officers

NEW YORK - The United States under President Donald Trump is ramping up use of surveillance systems and artificial intelligence (AI) to track and arrest immigrants, raising fears that risks to accuracy and privacy could put almost anyone in danger of getting caught up in the crackdown.

The Department of Homeland Security (DHS) and other immigration control agencies are using a suite of AI tools — such as facial recognition scanners in public areas and robotic dogs patrolling the southern border for human movement — as part of the crackdown on alleged illegal immigration.

Many of the AI tools that immigration agents are using have been in place for years and are a legacy of previous administrations, according to Saira Hussain, an attorney at the Electronic Frontier Foundation, a digital rights advocacy group.

But now these tools have "a ramped-up scope in terms of who (they are) targeting," and a wider range of people will have access to the data these tools collect, Hussain said.

The pumped-up surveillance dragnet also includes services run by private contractors like Babel Street, which trawl immigrants' social media accounts to collect personal information.

Once that information is collected, agencies like DHS and the U.S. Customs and Border Protection (CBP) use it to track locations of immigrants, map out their family trees and justify arrest warrants and deportation decisions.

An example of the widening scope is the government’s new “Catch and Revoke” program, launched under Secretary of State Marco Rubio in March.

It uses AI to monitor the public speech of foreign nationals, particularly student visa holders, to locate those “who appear to support Hamas or other designated terror groups,” the Axios news website reported.

Anyone caught by the programme is at immediate risk of losing their visa, and more than 300 foreign nationals, including those with student and visitor visas, have had their visas revoked under the initiative, according to Rubio.

"If they're taking activities that are counter to our ... national interest, to our foreign policy, we'll revoke the visa," Rubio told a news conference on March 28.

Accuracy risks

Digital rights advocates point to the tendency of AI tools to spout false "hallucinations" - answers or information that looks real but is fabricated – which make them dangerous to use in situations that require precision like immigration enforcement.

Paromita Shah, executive director at immigrant rights firm Just Futures Law, said the arrests of migrants using these tools "raise a lot of concerns about civil rights violations and abuses."

Since Trump took office in January, there have been numerous cases of immigration officials acting on inaccurate AI data, rights advocates say.

These include Jonathan Guerrero, a U.S. citizen arrested by U.S. Immigration and Customs Enforcement (ICE) agents in Philadelphia, and Jensy Machado, a U.S. citizen held at gunpoint while driving to work in Virginia. Each was later released.

An executive order signed by Trump in January suggested the possible return of "Rapid DNA testing," a process used to verify migrant family connections that was scrapped in 2023 due to privacy and accuracy concerns.

"Technologies start at the border and creep into the interior without being proven to be accurate," said Hussain.

"I think accuracy is not what this administration is going for. They're really going for the big splashy news of ‘We were able to take down X many people.'"

Tekendra Parmar, an independent tech analyst, agreed, saying accuracy is not a priority for the Trump administration that is more concerned with meeting deportation targets.

"The fallibility of the technology ... allows the current administration to create a rubber stamp deportation policy under the guise of artificial intelligence," Parmar wrote in Compiler.

Neither DHS nor ICE responded to requests for comment.

Casting a wider net 

Surveillance systems do not just target immigrants but rather all U.S. residents, citizens or not, researchers say.

In 2021, Georgetown University Law Center researchers found ICE had access to driver's license data of three in four U.S. adults and could locate the same number through their public utility records.

"These data intensive tools aggregate all of these data points and create associations," said Emerald Tse, an associate at Georgetown Law's Center on Privacy and Technology. "(They) can implicate the people in your household, your neighbours, your workplace, literally every aspect of your life."

That aggregated data is pumped into algorithms that help decide who ICE should detain, whether to release a person from detention or determine the terms of their electronic surveillance, experts say.

Immigration agencies are growing in reach as well. Another Trump executive order encourages the use of what are called 287(g) agreements that allow the DHS to deputize local law enforcement to act as federal immigration officers.

This gives local authorities full access to the AI tools that ICE uses, along with all the private data those tools have gathered. That means thousands more immigration agents are handling private data and hunting the people that data implicates.

"That's where I see a ramp up," said Hussain. "The feds have their technology, and locals have their technology. There'll be a lot of sharing of that information and whatever data the technology is able to collect."

(Reporting by William Antonelli; Editing by Anastasia Moloney and Ellen Wulfhorst.)