“It sounds like inside surveillance.” That is how one ICE official, speaking anonymously for fear of retaliation, described the agency’s new plan to hide its employees’ identities with artificial intelligence, according to The Intercept.
The Intercept reported on September 2, 2026, that ICE has launched a “Doxing Mitigation Initiative.” An email went to all agency employees on Monday, signed by Acting ICE Director David Venturella, and ICE employees leaked it to the outlet. The memo names the company behind the effort as ZeroFox and calls its product “a modernized protective technology solution designed to strengthen safeguards for our most valuable resource: You and your families.”
The contract
Public procurement records show that the Department of Homeland Security quietly signed a $15 million contract with ZeroFox in July for software licenses supporting ICE operations and investigations. The services are for the Office of Intelligence in ICE’s Homeland Security Investigations division.
ZeroFox builds AI software meant to stop data leaks across social media and the dark web. Here is what that means in plain terms: it hunts for information online, and in this case the target is anything that could identify, or “dox,” an ICE employee. The Intercept said neither the contract nor the internal email explains exactly how the program will work. ICE and ZeroFox did not respond to requests for comment.
The worry from inside
The anonymous official fears the technology could be turned on whistleblowers and critics of the Trump administration. “If things go sideways they may just throw all your info out there, make you out to be the problem or dox you,” the official said. “I don’t trust it.”
A second ICE official, also anonymous, said some personnel already use third-party services to keep their information private. “I’m not really keen on giving my PII to a company that doesn’t have a solid track record with government contracts,” that official said.
Richard Forno, associate director of the cybersecurity institute at the University of Maryland, Baltimore County, told The Intercept it is plausible the tech could be used to find whistleblowers. He sketched one method: leadership seeds a piece of information, it leaks to a reporter, and the trail leads back to the source. “You’re kind of setting up a honey pot situation to figure out who the leakers are,” Forno said.
Sophia Cope, a senior staff attorney at the Electronic Frontier Foundation, said it is understandable to want less personal information online. But she drew a line. If the program is meant “to shield officers and the agency itself from public accountability, especially when information is already publicly available,” she said, “that’s a serious concern.”
ZeroFox’s record
The company has a history with protest monitoring. In 2015 it offered Baltimore officials its services for free and sent police a list of “threat actors” that included peaceful Black Lives Matter organizers. Its co-founders told the press then that the labels likely reflected follower counts, and that “physical” meant people acting in real life rather than only online. After the January 6, 2021, attack on the Capitol, the Washington Post reported, FBI agents struggled with their ZeroFox technology and gave it a derisive nickname.
Haveli Investments, a private equity firm, bought ZeroFox in 2024 for roughly $350 million. Bryan Ware, a former assistant director for cybersecurity at DHS in the first Trump administration, is now an executive there.
