The team is said to communicate primarily via Signal, an encrypted messaging app that allows users to send disappearing messages
A secretive team of technologists operating under the Department of Government Efficiency (DOGE), reportedly led by Elon Musk, is facing scrutiny for allegedly deploying artificial intelligence (AI) tools to monitor internal communications within at least one federal agency. According to a Reuters investigation citing individuals familiar with the matter, the team is said to be searching for signs of disloyalty to US President Donald Trump.
Sources with direct knowledge of the situation told Reuters that employees at the Environmental Protection Agency (EPA) were informally warned that their emails and virtual messages may be under surveillance. The AI tools reportedly search for language considered hostile towards Trump or Musk. The move is seen as part of a broader effort by DOGE to overhaul bureaucratic processes, reduce perceived inefficiencies, and eliminate internal dissent.
While DOGE’s official mandate is to streamline government operations, its methods — which allegedly include the use of Musk’s AI chatbot Grok and encrypted messaging platforms — are raising concerns among cybersecurity professionals and government transparency advocates.
The team is said to communicate primarily via Signal, an encrypted messaging app that allows users to send disappearing messages. This practice has raised legal red flags, particularly around federal record-keeping. “If they’re using Signal and not backing up every message to federal files, then they are acting unlawfully,” said Kathleen Clark, a government ethics specialist at Washington University in St. Louis.
Further complicating matters, DOGE staff are reportedly avoiding traditional documentation protocols by editing official drafts collaboratively in Google Docs — a tactic that, according to insiders, helps them make swift changes while leaving little trace.
At the EPA, where the agency has already been affected by budget cuts and staff reductions, employees were reportedly cautioned: “Be careful what you say, what you type and what you do.” While the agency acknowledged that it is “looking at AI to better optimise agency functions,” it denied that such tools are influencing personnel decisions.
Federal court records cited by Reuters also reveal that DOGE has restricted access to several government systems and databases. At the Office of Personnel Management (OPM), which oversees federal hiring and HR policy, over 100 technology staff reportedly lost access to cloud-based systems earlier this year. Only two people — a Trump-appointed Chief Information Officer and one career official — now have administrative control over sensitive personnel data.
The DOGE initiative is currently facing legal scrutiny. In March, a judge ordered the release of internal documents following a lawsuit filed by Citizens for Responsibility and Ethics in Washington (CREW). The watchdog group has raised serious concerns about the team’s secrecy, operational practices, and refusal to adhere to transparency requirements under federal law. As of this week, the documents have not yet been made public.
The growing influence of AI in federal operations, especially in sensitive areas like employee surveillance, is prompting renewed debate over privacy, accountability, and the balance between innovation and civil liberties in government systems. Legal experts and digital rights advocates are now calling for greater oversight of DOGE and its use of emerging technologies.
As questions mount, both the ethical and legal implications of DOGE’s work — particularly its reported monitoring of internal government communications — are likely to remain in sharp focus in the weeks ahead.