In an astonishing disclosure, sitting at the crossroads of technology, politics, and governance, sources hold that Elon Musk’s somewhat Department of Government Efficiency (DOGE), is said to be employing artificial intelligence in order to monitor communications within at least one U.S federal agency. The alleged aim is to detect and eliminate those thoughts that have harmful inclinations toward President Donald Trump or his agenda.
While the majority of DOGE’s internal functioning remains covered in secrecy, two sources familiar with the situation say that this use of AI constitutes an unprecedented type of political surveillance. The advanced technologies are being deployed to examine the loyalty of federal workers working in a situation characterized by massive lay-offs and huge budget cuts.
Further, the DOGE team members are also said to be exchanging messages using the encrypted Signal messaging application, which raises red flags in terms of legal and ethical implications. The app provides for self-destruction of messages within a specified time as well. Kathleen Clark, who specializes in government ethics, states that such an activity is not just suspicious legally but will also be extremely dangerous from an abuse of government power standpoint. She said,
“If they’re using Signal and not backing up every message to federal files, then they are acting unlawfully.”
“It sounds like an abuse of government power to suppress or deter speech that the president of the United States doesn’t like.”
Gemini and the AI Purge
Musk’s very own AI technology is at the heart of the operations in question. Sources say that its DOGE has been “heavily” utilizing Grok, Musk’s response to ChatGPT, although sources could not confirm the nature of this deployment. Meanwhile, another advanced AI model, Gemini, is claimed to be helping with the analyzing of communications, especially those from the likes of Environmental Protection Agency (EPA), for anti-Trump or anti-Musk sentiment. A source familiar with the EPA said,
“We have been told they are looking for anti-Trump or anti-Musk language.”
The EPA, which is already under the cannon by this administration, has had nearly 600 employees placed on leave this year and plans to trim down its budget by 65%. It’s alleged that Trump appointed officials warned EPA management that DOGE AI is being utilized to monitor internal platforms such as Microsoft Teams, instructing staff to
“Be careful what you say, what you type and what you do.”
In a vague statement, the EPA noted that it was considering AI to enhance operations, but it did not state that AI had played any role in personnel matters. EPA acknowledged in a statement that it was
“looking at AI to better optimize agency functions and administrative efficiencies. It was not using AI, as it makes personnel decisions in concert with DOGE.”
The agency did not specifically address the allegations of personnel surveillance.
Government of Ghosts
Across the political spectrum, different political voices are criticizing DOGE for what seems like quiet engineering bloodshed. The concern is political loyalty over competence, and the fear is that phantom bureaucrats, a group of engineers and AI systems, are rewriting the rules of governance. With each alleged illegality, court ordered DOGE to operate under federal transparency laws in compliance with a CREW lawsuit involving the use of disappearing messages and lack of document tracing accountability. So far, DOGE has not produced a single record it was ordered to furnish henceforth.
Meanwhile, it has also been reported that DOGE stripped hundreds of tech personnel from cloud systems on the premises of key government agencies, including the Office of Personnel Management (OPM), containing the personal data of millions of federal employees. Only two people have access to it, one of which is Greg Hogan, a Musk-friendly political appointee with some AI startup experience.
Now, with an opinion of a brutal takeover of operations, DOGE is riding the fast lane and skipping the radar of career public servants with unprecedented speed. The source said,
“The concept was that through taking the government data they could build the most dynamic AI system ever. AI could then do the work.”
It is breaking convention by employing collaborative tools like Google Docs as opposed to bureaucratic processes, leaving career civil servants in the dark. The source said, referring to the online word processing software, “There’s multiple people in one Google Doc editing things simultaneously.”
Broader Scenario
When transparency is discarded, and AI is used not merely as a tool but rather as a custodian of an ideology, governance approaches on becoming a form of surveillance. While Musk may envision saving the government from itself, it rather seems to be preventing it from democracy. The question henceforward should not be whether AI can reduce bureaucracy and make it more efficient, it is really about whether we are willing to let it decide who stays and who goes.
In a more general sense, whether you are for or against Musk, he seems to be raising the most intense question on other governments. Who gets to write the algorithms that run a nation? And what happens when loyalty to a political leader is programmed into code? For now, all these questions remain hidden within a veil of secrecy, encrypted chats, and AI-driven files. The storm clouds are definitely gathering and so are the lawsuits.
Author