The Trump-Vance administration has rapidly rolled out artificial intelligence (AI) systems throughout the federal government while dismantling guardrails meant to ensure that AI is safe, secure, and operating within the law. Elon Musk and his so-called “Department of Government Efficiency” (DOGE) are overseeing the deployment of these opaque systems. Some of the most concerning uses of AI reported in the media include the administration’s attempts to surveil federal employees to try to determine their productivity or ideological opinions. This information could potentially be used for personnel decisions, such as termination of employment or demotion. AI systems are known for their biases and production of incorrect results, meaning it is highly unlikely these results would be objective or reliable enough for making such critical decisions.

Government employees have always known that their online activities are monitored for records retention, counterintelligence, and other reasons. But what is new is how the administration could use the data.

President Trump’s history of retaliating against perceived enemies, undermining First Amendment rights, and attempting to dismantle the federal workforce raises serious concerns about the unknown scope and intent of the surveillance targeting government employees.

There has been no indication from this administration about how collected records will be analyzed, and what will be done if the administration decides that workers are speaking or acting in ways it does not like.

Democracy Forward is tracking the deployment of AI across the federal government, including at:

  • Department of State: IT staff at the agency have said that monitoring software has been installed on government computers, including software that records individual keystrokes, which employees anticipate will be fed into AI tools. Monitoring State Department employees for their loyalty to President Trump would align with his February executive order calling for any foreign service officer who dissents from or contradicts President Trump’s foreign policy to be fired.
  • General Services Administration (GSA): A Trump political appointee at GSA has openly called for the agency to adopt an “AI-first strategy.” DOGE staffers have reportedly told GSA’s managers that “they plan to automate a majority of jobs” and “‘[replace] the human workforce with machines.” GSA has also rolled out a chatbot to automate the work previously done by employees who have been fired since the beginning of the Trump administration. Additionally, reporting indicates that DOGE employees are working on a second chatbot that could “plug into” GSA’s vast stores of government data.
  • Environmental Protection Agency (EPA): Meetings and phone calls held with EPA staff may be secretly recorded by AI tools without notifying call participants. DOGE staff are also reportedly using AI to scan employees’ emails and private messages for any signs of statements that could be considered “anti-Musk, anti-Doge, or anti-Trump.”
  • Department of Veterans Affairs: Senior officials warned employees that their virtual meetings are being secretly recorded, and that they should be careful if voicing any opinions that could be considered counter to the president. DOGE staff have also introduced an AI tool to write code for the agency, which would require it to access sensitive data on contracts and veterans. This effort is being spearheaded by a DOGE employee who has called for AI to replace humans in many workplace jobs.

We are also monitoring the ways in which the administration may be using, or attempting to use AI, to:

  • Revoke grants and contracts that reference words related to diversity, equity, inclusion, or accessibility;
  • Surveil and cancel the visas of immigrants exercising their First Amendment rights;
  • Rewrite government rules and regulations to align them with a far-right agenda; and
  • Analyze vast swaths of information held by agencies such as the Social Security Administration, which risks exposing Americans’ private and sensitive information to fraud.

As part of our investigations, we are seeking access to public records regarding the administration’s use of AI. To date, we have sent dozens of records requests to more than ten government agencies asking for the following types of documents:

  • Top agency officials’ communications with DOGE employees
  • Top agency officials’ communications including key terms related to AI
  • Agency validation studies and privacy impact assessments on the use of AI for personnel matters
  • Agency guidance related to the use of AI for personnel matters
  • Source code, training materials, or parameters given to AI systems deployed for personnel matters
  • Searches, instructions, queries, or inputs into AI systems deployed for personnel matters
  • Outputs, recommendations, or reports created by AI systems deployed for personnel matters

We are waiting for a meaningful response to our FOIA requests from the government.