One of the key issues highlighted in the report is the team’s reliance on Signal, an encrypted messaging app that allows messages to vanish after being read
A covert team of technologists led by Elon Musk, operating under a little-known government unit named the Department of Government Efficiency (DOGE), is reportedly using artificial intelligence to monitor internal communications at a federal agency for perceived disloyalty to President Donald Trump. The development, first reported by Reuters, has raised concerns among cybersecurity experts and legal observers.
According to individuals with direct knowledge of the matter, employees at the Environmental Protection Agency (EPA) have been cautioned that AI tools may be reviewing emails and virtual chats for any language viewed as critical of Trump or Musk. Insiders say this surveillance effort is part of a broader initiative by DOGE to streamline federal operations, suppress dissent, and overhaul bureaucratic processes.
While DOGE’s official mandate focuses on cutting waste and inefficiency, its methods have attracted scrutiny. The team reportedly uses encrypted communication channels and Musk’s AI chatbot, Grok, as part of its operations. Experts argue that these tactics, though high-tech, raise serious ethical and legal concerns.
One of the key issues highlighted in the report is the team’s reliance on Signal, an encrypted messaging app that allows messages to vanish after being read. Kathleen Clark, a government ethics specialist from Washington University in St. Louis, warned that such practices may breach federal record-keeping laws. “If they’re using Signal and not backing up every message to federal files, then they are acting unlawfully,” she said.
Additionally, DOGE staff have been accused of bypassing traditional document handling procedures by using platforms like Google Docs to collaboratively edit drafts in real time, thus avoiding the creation of permanent records. Insiders claim this approach enables quick changes to government processes without leaving a trace.
Within the EPA, staff were reportedly warned to “be careful what you say, what you type and what you do,” following the rollout of AI monitoring tools. The agency, which has faced funding cuts and staffing reductions under Trump’s administration, confirmed it is exploring AI for operational efficiency, but denied using it for monitoring or personnel decisions.
Further complications emerged when federal court records revealed DOGE had restricted access to essential government systems and data. At the Office of Personnel Management (OPM), over 100 technical staff reportedly lost access to cloud systems earlier this year. Administrative control was said to be limited to just two individuals — one career employee and a Trump-appointed Chief Information Officer.
The activities of DOGE have not gone unchallenged. In March, watchdog group Citizens for Responsibility and Ethics in Washington (CREW) filed a lawsuit, prompting a judge to order the release of internal records. The group cited DOGE’s unusual secrecy and non-compliance with transparency requirements. However, as of this week, no documents have been disclosed.
As scrutiny grows, questions remain about the scope of DOGE’s mandate, the legality of its practices, and the broader implications of using AI to monitor government employees. The episode underscores the complex tensions between technology, governance, and accountability in an increasingly digital federal workforce.
