In the weeks after Donald Trump’s 2024 Presidential victory, Joshua Aaron, a software developer in Texas, kept asking himself, “What can I possibly do to help?” On the campaign trail, Trump had vilified immigrants, promising to carry out “the largest deportation operation in the history of our country.” Aaron, the son of a rabbi, had grown up around Holocaust survivors and saw parallels to the rise of the Third Reich. “They’re running the same exact playbook,” he told me. “Making people afraid of their neighbor, militarizing the streets, all under the guise of ‘We’re going to keep you safer.’ Safer from whom?” Aaron thought about writing an op-ed or speaking at a rally, but such gestures struck him as painfully insufficient. He wanted to give power to other people, so he thought of what’s most often in their hands. “We take our phones everywhere we go,” he said. “We take it into the bathroom with us. We take it to the shopping center with us. It’s in the car with us. Having that with you at all times gives you a sense of security.”

Aaron decided to develop ICEBlock, an app that allowed users to report and view ICE sightings within a five-mile radius of their location. Posts expired after four hours, because he figured that agents were unlikely to stay in one place for long. The platform was designed to act as an early-warning system, much like the “police reported ahead” feature in the navigation app Waze. “If you’re walking down the street and four blocks ahead of you somebody reports a sighting, turn left, turn around, go home,” he explained. “Just avoid that confrontation in the first place.”

When Aaron submitted the app to Apple, last February, it was initially rejected. “They wanted to vet it,” he said. “You know, Is this legal?” The fact that he had no plan to monetize the application—ICEBlock would not track users’ location or collect any of their data—seemed to raise suspicions. But after multiple meetings between Aaron and Apple’s representatives, the company seemed satisfied that, per its guidelines, ICEBlock would not solicit or encourage criminal behavior. The app was released last April, to little fanfare. Around three thousand people had downloaded it by June 30th, when CNN ran an interview with Aaron. “Then it just kept growing and growing,” he told me. By the next day, ICEBlock was the most downloaded free social-networking app in the App Store.

Aaron said that his inbox was flooded with e-mails from grateful users. Mothers living abroad were relieved there was an app to check if ICE agents were near their children, and the heads of foreign companies thanked Aaron for keeping their U.S. employees safe. Social-media posts show that ICEBlock was used across the country, from the Bronx to Opelika, Alabama, with sightings of federal agents reported at car washes, Mexican restaurants, and grocery stores. Some immigrant advocates worried that there was no mechanism to verify that the information posted on the application was real. (“Every time we would go and respond, it was not ICE,” a volunteer at the mutual-aid group NorCal Resist told me. “It was police or a smog check or a D.U.I. checkpoint.”) Aaron said that incorporating a method of moderation or verification—such as the voting system in Waze—would have required collecting user data that might someday fall under a subpoena. “What’s the worst-case scenario?” he asked me. “I opted for ‘Stay away for a couple of hours,’ because nobody gets hurt.”

Even if ICEBlock wasn’t always accurate, the Trump Administration was quick to view it as a threat to its mass-deportation campaign. Shortly after the CNN interview aired, Kristi Noem, who was then the Homeland Security Secretary, posted on X, “This sure looks like obstruction of justice.” ICE’s acting director, Todd Lyons, accused CNN of “enabling dangerous criminal aliens to evade U.S. law” and “willfully endangering the lives of officers.” The following morning, the Administration’s border czar, Tom Homan, said that he was “begging the Department of Justice to look into this and hold people accountable.” Pam Bondi, the Attorney General, soon confirmed that her department was taking aim at Aaron: “We are looking at it, we are looking at him, and he better watch out, because that’s not a protected speech.”

Aaron was dismissive of the Administration’s threats. “It couldn’t have been better publicity, honestly,” he said. “Let them talk.” ICEBlock included a disclaimer that it was intended for “information and notification purposes only” and was “not to be used for the purposes of inciting violence or interfering with law enforcement.” Such measures may have staved off prosecution, but they haven’t prevented other kinds of attacks. Right-wing activists soon learned that Aaron’s wife, Carolyn Feinstein, was an auditor at the Justice Department. Andrew Wilson, an avowed paleoconservative who hosts an online debate show called “The Crucible,” shared on X the couple’s address and phone numbers. Trump’s self-proclaimed “loyalty enforcer,” Laura Loomer, chimed in, too. “This is yet another vetting failure by Pam Blondi,” she posted. “Carolyn Feinstein needs to be FIRED from the DOJ immediately.”

Two days later, on July 18th, Feinstein received a termination notice. The letter cited “lack of candor,” even though she had informed the government that she was married to Aaron and held a share in the company that ICEBlock was registered to (a precaution in case something happened to her husband). Aaron told me that Feinstein had played no role in the development of the app and had never downloaded it. He described her as scrupulous and hardworking. “They couldn’t get to me,” Aaron said, “so they got to my wife.”

A week later, a deliveryman showed up at the couple’s home with a pizza. The name on the order was Theodora Kaczynski—a female version of the Unabomber. “It was an obvious ‘We know where you live’ kind of thing,” Aaron said. Soon, the couple was receiving deliveries from slice shops across the city. “It was an entire day of just pizza after pizza after pizza.”

Meanwhile, the Trump Administration’s expansion of immigration enforcement inside American cities brought even more attention to Aaron’s app. By the end of August, ICE and C.B.P. officers in Los Angeles arrested more than five thousand immigrants. ICEBlock reached a million users that month, with more than ten thousand joining each day. Aaron spent long hours at his computer answering queries, fixing bugs, and releasing updates to serve his growing base. He was also working on a version for Android, the operating system immigrants tend to use. “Whenever this Administration says ‘ICE-tracking apps,’ we know they’re talking about ICEBlock,” he told me. “They just stopped using the name, because they realized that, the more they do, the bigger it gets.”

Before sunrise on September 24th, a twenty-nine-year-old gunman with a bolt-action rifle climbed onto the roof of an immigration attorney’s office in Dallas and began firing rounds at the city’s ICE headquarters, a two-story building across the street. Bullets sprayed its walls and windows and pierced a van carrying a group of shackled immigrants. Norlan Guzmán Fuentes, a landscaper from El Salvador, was fatally wounded. Miguel Ángel García, a housepainter from Mexico, was taken to a hospital, where he died five days later. His wife was about to give birth to their third child.

No law-enforcement agents were hurt, but the shooter, Joshua Jahn—who was found dead at the scene—had left behind notes indicating that the officers had been his intended targets. “Hopefully this will give ICE agents real terror,” one of the notes said. “To think, ‘is there a sniper with AP rounds on that roof?’ ” An unspent shell casing read, in dark-blue letters, “ANTI-ICE.”

Even though the shooting had occurred at a government building with a known address, officials began to claim that ICE-tracking apps had enabled Jahn’s actions. The day after the attack, Kash Patel, the F.B.I. director, wrote on X that Jahn had searched for apps that shared the locations of ICE agents. Marcos Charles, an ICE associate director, claimed that Jahn had actually used those apps, and blamed their creators and distributors for putting agents in danger. “It’s a casting call to invite bad actors to attack law-enforcement officers,” he said at a press conference that afternoon. “It’s no different than giving a hit man the location of their intended target.”

On September 30th, a few days after the attack, Loomer called out Tim Cook and Sundar Pichai, the heads of Apple and Google, for making such programs available. “The shooter who shot up ICE agents in Dallas, Texas was using an ICE tracking app,” she wrote on X. “Why are you going to very fancy dinners at the White House and kissing President Trump’s ass while also allowing for this type of lawlessness and criminal activity to take place in your App Stores?”

Two days later, Apple took down ICEBlock. A message on Aaron’s developer portal said that the company had received information from law enforcement which showed the app’s purpose was “to provide location information about law enforcement officers that can be used to harm such officers individually or as a group.” Bondi was quick to take credit. She told Fox News that the Justice Department had “reached out to Apple today demanding they remove the ICEBlock app from their App Store—and Apple did so.”

At least four similar apps were axed in the following days, including Eyes Up, a platform that archived videos of arrests, raids, and abuses by immigration agents. Its creator, Mark Hodges, received the same message as Aaron, even though Eyes Up did not provide agents’ real-time locations. In both cases, Apple cited a rule that forbids discriminatory, defamatory, and mean-spirited content directed at “targeted groups” such as racial, religious, and sexual minorities.

Around the same time, Google removed from its Android store at least three apps that tracked immigration operations. The company told 404 Media that the apps were taken down because they publicized the location of a “vulnerable group” that had recently faced a violent attack. A couple of weeks later, Meta removed a Facebook page that published ICE sightings in Chicago. A company spokesman said the page had violated the platform’s policy on coördinated harm, which bans “outing the undercover status of law enforcement.”

The Administration’s treatment of immigration officers as a vulnerable group rests on claims that agents now face unprecedented threats. The day after the Dallas shooting, Trump issued a Presidential memorandum called “Countering Domestic Terrorism and Organized Political Violence,” which cited a thousand-per-cent increase in attacks on ICE officers and claimed that such acts of violence were not “a series of isolated incidents” but the “culmination of sophisticated, organized campaigns of targeted intimidation.” The President called for a “new law enforcement strategy that investigates all participants in these criminal and terroristic conspiracies” and instructed the Justice Department to prosecute anyone involved to “the maximum extent permissible by law.” In a subsequent memorandum to federal prosecutors, which was leaked to the press, Bondi wrote that “rampant criminal conduct rising to the level of domestic terrorism” included “organized doxing of law enforcement.”

In November, after months of refusing to release hard numbers, D.H.S. reported that there had been two hundred and thirty-eight assaults on ICE officers between January 21st and November 21st, up from nineteen during the same period the previous year. A number of factors likely contributed to the increase. ICE announced that it hired twelve thousand officers last year, more than doubling the department’s operational workforce. The agency waived age limits for new recruits, relaxed its vetting process, and decreased the training period for incoming officers. Its agents were instructed to shift their focus from immigrants with criminal records to anyone who might be in the country illegally. By mid-October, ICE had carried out around a hundred thousand “at-large” arrests—in homes and public spaces, at worksites—more than in any year of the first Trump Administration.

Even so, when reporters at the L.A. Times analyzed court records in Chicago, San Diego, Portland, Los Angeles, and D.C.—cities where the Trump Administration had conducted mass raids or deployed the National Guard—they found only a twenty-six-per-cent increase in complaints of attacks against federal law enforcement. ICE and Border Patrol agents were presumed the victims in about sixty per cent of cases. Most of the incidents had not resulted in physical injuries; in some cases, prosecutors charged people for assaulting agents with such weapons as a tambourine, an umbrella, and a Subway sandwich. Multiple defendants, on the other hand, appeared to have been hurt in the altercations. In Los Angeles, Jonathon Redondo-Rosales was charged with assault for smacking an agent with a hat after a car struck him. In Chicago, Marimar Martinez was accused of ramming her car into a Border Patrol vehicle, despite video evidence suggesting the officer swerved his car into hers; an agent subsequently shot her five times. Since Trump returned to the White House, more than forty people have died in encounters with ICE or while in the agency’s custody, but no deportation officers have been killed on duty since ICE was created, in 2003.

Documents obtained through open-records requests by Property of the People, a transparency nonprofit, show that the Trump Administration has made a concerted effort to warn immigration officers that they are facing heightened risks. “Everyone from federal agents to beat cops have been inundated with dubious safety alerts,” Ryan Shapiro, the organization’s director, told me. He sent me more than a dozen unclassified documents about ICE-tracking apps, many of them created by fusion centers—intelligence hubs created after 9/11 to share information about criminal and terrorist threats. Some of the earliest reports focus on People Over Papers, a map that allowed users to anonymously submit ICE sightings. Unlike reports on ICEBlock, posts tended to feature detailed information about each incident, including photographs and videos. Last February, the Vermont Intelligence Center concluded that “malicious actors seeking to cause harm to law enforcement and immigration officers” could use initiatives such as People Over Papers to carry out attacks.

The first government documents that Shapiro obtained about ICEBlock attempted to establish a connection between the application and assaults on immigration agents. On July 23rd, the F.B.I. issued a report about ICEBlock that cited two recent attacks against D.H.S. facilities in Texas which the department had attributed to violent extremists. During a Fourth of July protest outside an Alvarado detention center, at which demonstrators were setting off fireworks, an assailant fired a gun, wounding a police officer. A few days later, a young man shot at a Border Patrol building in McAllen, injuring one employee and two officers. The Bureau’s D.C. office lacked “specific reporting regarding the current use of ‘ICEBlock,’ ” but it was concerned about “the combination of rising violence and capabilities of the application.”

Scarlet Kim, a senior attorney at the A.C.L.U. who reviewed the documents, told me that the connection was spurious. “The government is not substantiating those concerns of safety threats with actual threats,” she said. “They point here and there to attacks on facilities, but they have in no way connected the apps to those attacks.”

Intelligence and law-enforcement agencies have also repeatedly raised concerns that tracking apps could be used to dox immigration agents—which, according to a D.H.S. analysis from last June, carried the “downstream threat of violence by violent opportunists or domestic violent extremists.” Traditionally, doxing refers to the release of someone’s personal information—a home address or phone number—with the intent to intimidate or to encourage others to retaliate. Facts about law-enforcement officers, such as their names, badge numbers, and work license plates, don’t typically qualify. Nor does the location of on-duty agents, or recordings of them in public spaces. Even if sharing such information could be considered doxing, Kim said, it would be protected by the First Amendment, unless it fell within a narrow exception, such as an incitement to violence. “These documents suggest that government officials should be particularly immune from doxing, but that’s just not true,” she told me. “To the contrary, the power that officials exercise makes it even more important that people be free to publish information about them.”

Yet immigration agents now routinely wear masks, drive unmarked cars, and refuse to identify themselves. At the same time, they use advanced technology to profile migrants and protesters, including an app built by Palantir that helps to locate potential deportation targets and to determine which neighborhoods to raid.

“Agency responses to ICE-tracking apps are illustrative of a broader drive to target even the most mundane forms of anti-ICE organizing and pro-immigrant sentiment as extremism and terrorism,” Shapiro told me. An F.B.I. report from November shows that the agency was pursuing dozens of investigations into perceived threats against immigration enforcement, more than a quarter of which were classified as domestic terrorism. Separately, it was conducting NSPM-7 domestic-terrorism investigations in twenty-seven cities. When I asked Shapiro if he foresaw the government opening terrorism investigations into developers such as Aaron, he said, “It’s a near-certainty that it’s already happening.”

Some activists have found creative ways to continue providing a platform for reports on ICE. People Over Papers, which was taken down from the platform Padlet following pressure from Loomer, now functions as a progressive web app—users can add it to their home screens and receive notifications as they would on a normal application, without having to go through an app store. More than five hundred thousand people have signed up. Other platforms, such as Eyes Up, are still accessible as traditional websites. The government, in turn, is searching for new ways to shut these services down. “Some of the websites have publicly available source code under an open-source license, allowing them to easily be replicated or copied,” another F.B.I. report from November warned. The Bureau called for greater collaboration with technology companies to “remove websites or applications that pose a public-safety risk.”

ICEBlock, on the other hand, is now available only to users who had downloaded it before it was deleted by Apple. Aaron told me that the program’s design and large user base have made it all but impossible to turn it into a progressive web app. “I would rather fight it out in the courts in the hopes that we can get it reinstated and declared protected speech, rather than try to bring a half-assed solution to the market that will end up hurting people,” he said.

On December 8th, Aaron filed a lawsuit against Bondi, Homan, Noem, and Lyons. The complaint alleges that the pressure they put on Apple to remove ICEBlock violated the First Amendment, which prohibits officials from “coercing private entities to suppress disfavored expression.” It also holds that their threats against Aaron were designed to chill his speech and to deter others, such as CNN, from amplifying or facilitating it. The federal district court of D.C. has been asked to declare those actions illegal and to forbid government officials from threatening, investigating, or prosecuting Aaron and from pressuring Apple or any other distributor of ICEBlock to block the application.

The case could have wide-ranging consequences on efforts to document and track the activities and abuses of immigration agents. If Aaron is successful, the court will set a precedent that may not only protect him but also shield other activists and developers from censorship and prosecution. On February 11th, Hodges, the creator of Eyes Up, along with the founder of the Chicago-sightings Facebook page similarly sued Bondi and Noem for violating their First Amendment rights.

Earlier this month, the government requested that Aaron’s lawsuit be dismissed, largely on the ground that he had failed to show that officials had coerced Apple into taking down his app. A judge has yet to decide whether the case will proceed. “I can’t control what they are going to try to do,” Aaron told me. “I will fight to my death against it, but I will not live in fear. If you’re afraid, they win.” ♦