DOGE Cuts 222,000 Jobs, AI Surveillance Raises Privacy Concerns

The Department of Government Efficiency (DOGE) has implemented significant job cuts, totaling around 222,000 in March alone. These reductions are particularly impactful in areas crucial to maintaining America's competitive edge, such as artificial intelligence (AI) and semiconductor development. The cuts have raised concerns about the potential sabotage of the U.S.'s technological advancements, as these sectors are pivotal for future innovation and economic growth.

Beyond the workforce reductions, there are growing concerns about the use of AI for surveillance within federal agencies. Reports indicate that DOGE is employing AI tools to monitor federal employees' communications, searching for signs of disloyalty. This practice has already been observed within the Environmental Protection Agency (EPA) and is part of a broader plan for extensive government cuts. Federal workers, accustomed to email transparency due to public records laws, now face the scrutiny of hyper-intelligent tools analyzing their every word.

The deployment of AI for surveillance raises critical questions about trust and privacy. The use of AI in complex bureaucracies can introduce biases and other issues, as highlighted by the General Services Administration's (GSA) own help page. The increasing consolidation of information within AI models poses a significant threat to privacy, and there are concerns that DOGE's actions may violate the Privacy Act of 1974. This act, enacted during the Watergate scandal, aims to prevent the misuse of government-held data. The act stipulates that no one, including special government employees, should access agency "systems of records" without proper authorization. DOGE's actions, under the guise of efficiency, may be jeopardizing Americans' privacy.

Surveillance in the modern context extends beyond cameras and keywords; it involves who processes the signals, who owns the models, and who decides what matters. Without strong public governance, this direction could lead to corporate-controlled infrastructure shaping government operations. Public trust in AI will weaken if decisions are perceived to be made by opaque systems outside democratic control. The federal government is supposed to set standards, not outsource them.

The National Science Foundation (NSF) has recently cut more than 150 employees, with internal reports suggesting even deeper cuts are forthcoming. The NSF funds critical AI and semiconductor research across universities and public institutions, supporting everything from foundational machine learning models to chip architecture innovation. The White House is also proposing a two-thirds budget cut to the NSF, which could severely impact American competitiveness in AI. Similarly, the National Institute of Standards and Technology (NIST) is facing the loss of nearly 500 employees, including teams responsible for the CHIPS Act's incentive programs and R&D strategies. NIST runs the U.S. AI Safety Institute and created the AI Risk Management Framework, both of which are essential for maintaining America's technological edge.

DOGE's involvement also raises concerns about confidentiality. The department has gained sweeping access to federal records and agency datasets, with AI tools combing through this data to identify functions for automation. This move shifts public data into private hands without clear policy guardrails, opening the door to biased or inaccurate systems making decisions that affect real lives. There is a lack of transparency around what data DOGE uses, which models it deploys, or how agencies validate the outputs. Federal workers are being terminated based on AI recommendations, and the logic, weightings, and assumptions of those models are not available to the public, highlighting a governance failure.

Surveillance does not equate to government efficiency. Without rules, oversight, or basic transparency, it breeds fear. When AI is used to monitor loyalty or flag words like "diversity," it erodes trust in the government rather than streamlining it. Federal workers should not have to worry about being watched for doing their jobs or saying the wrong thing in a meeting. This situation underscores the need for better, more reliable AI models that can meet the specific challenges and standards required in public service.