In April, the General Services Administration announced plans to automate 1 million work hours annually after cutting nearly 40% of its staff since October 2024, with similar reductions being seen across the government workforce.

While the Elon Musk-led Department of Government Efficiency (DOGE) may have receded as a formal initiative, it has been hiring staff members who have been working across several agencies and accelerating further government automation.

Washington first adopted large-scale automation during World War II to manage massive military datasets, before its expansion into the postwar administrative state. Unlike previous waves, however, AI-driven automation is reducing jobs across both government and private industry without creating comparable replacement roles.

These systems are already shaping core government functions tied to state authority and legitimacy, including the use of military force. Reports on the Pentagon's Maven Smart System, deployed in the 2026 Iran conflict, offer a glimpse into how far the use of such technologies has advanced.

Launched in 2017, Maven is a network of contractor-built systems led by Palantir Technologies, with involvement from companies like Microsoft and Amazon. It integrates satellite imagery, drone feeds, radar and infrared sensors, and signals intelligence, along with dozens of other data sources. Computer vision algorithms, which have been trained on vast image datasets, classify the "battlefield objects" with an "AI Asset Tasking Recommender" suggesting strike options.

Two decades ago, this task took thousands of personnel to complete, but it can now be done by a handful of operators in seconds. Targeting output increased from fewer than 100 before Maven to more than 5,000 per day during the Iran war, said a National Geospatial-Intelligence Agency official to Wired.

Earlier versions of Maven have been used in Afghanistan, Ukraine, Iraq, Syria, Yemen, and during the seizureof Nicolas Maduro in Venezuela, and the technology has continued to evolve during the Iran conflict. While not fully autonomous, it is another step toward true agentic AI warfare, where AI systems move beyond assisting human decisions through automation toward identifying and carrying out tasks with minimal human input.

The Pentagon has sought $54 billion as part of its 2027 budget to move toward an "autonomous and remotely operated systems across air, land, and above and below the sea, including the 'Drone Dominance program.'"

It is the latest signal of Washington's intention to reduce human involvement in war, as troop numbers continue their decades-long decline, reducing by 64% between 1968 and 2025. Azerbaijan's use of loitering drones in Armenia in 2020 and Israel's use of AI-assisted warfare in Gaza show how easily countries can adapt to these systems. Russian and Chinese efforts to increase their autonomous systems capacity are already competing or outpacing those of Washington.