WASHINGTON — A leaked internal memo reveals that U.S. Immigration and Customs Enforcement is contracting with a controversial artificial intelligence firm to build technology that masks the identities of its agents in public records, a move the agency frames as protection against doxing. Privacy advocates and whistleblower protection groups are sounding alarms that the same system could be repurposed to monitor internal communications and trace leaks back to employees.
Here’s the tension in plain terms: the tool that scrubs an agent’s name off a public database is, at its core, an identity-tracking system. ICE says it wants the tool to keep its people safe from online harassment.
Critics worry the agency could use the same technology to identify whistleblowers within the agency. The memo, circulated to ICE employees by a senior official whose name has not been released, does not disclose the AI firm’s identity. The company is described only as controversial, with prior work with government agencies and ethical questions surrounding its technology.
No contract amounts, financial figures, or implementation dates appear in the document. ICE officials say the anonymization tool will reduce the risk of agents’ personal information being exposed online — a practice known as doxing.
The practical upshot for an agent is straightforward: less of their private life sitting in searchable government files. But the counterargument, laid out by critics, is that the same technology could be repurposed to monitor internal communications and trace leaks back to employees, thereby chilling whistleblowing activities. That prospect has chilling implications for whistleblowing, which relies on anonymity as its core protection.
If federal employees believe the agency can trace any unauthorized disclosure back to them with AI-assisted precision, the argument goes, they will think twice before surfacing misconduct. The U.S. Office of Special Counsel, which handles whistleblower complaints, has not commented on the matter. ICE has not issued a public response to the leak as of this writing.
What to watch next: whether ICE confirms the contract and names the vendor, and whether further details emerge as ICE and the AI firm respond to the leak. The broader question — how far federal agencies can push AI-powered surveillance before it collides with civil liberties — is not going away, and this leak just put a very concrete face on it.
