Industry Insights

The Pentagon Has 100,000 AI Agents — And They’re Already Replacing Office Work

Written by Fed Gov Today | May 7, 2026 3:57:02 PM

Original Broadcast Date: 05/10/2026

Presented by Workday Government

Artificial intelligence is rapidly changing how federal employees work, and according to Matthew Cornelius, the transformation is already happening at scale. During an interview on Fed Gov Today, Cornelius, managing director for Federal Industry at Workday Government, explains how agencies are moving beyond AI experimentation and beginning to integrate autonomous systems directly into daily operations.

Cornelius points to the Pentagon’s deployment of more than 100,000 AI agents through the Department of Defense’s GenAI.mil platform as a major indicator of how quickly the federal landscape is evolving. These AI agents are already supporting mission activities such as drafting after-action reports, analyzing operational data and reviewing images. For Cornelius, the significance is not simply the number of agents being used, but the fact that agencies are beginning to trust AI systems with meaningful operational responsibilities.

He compares the current moment to his own experience working at the General Services Administration years earlier, when teams of contractors manually followed leaders from meeting to meeting preparing PowerPoint slides and documenting notes. Today, he says, autonomous AI agents can perform many of those same administrative functions in far less time. The shift reflects a broader evolution in how government organizations think about work itself.

Cornelius argues that agencies must now start treating AI agents much like they treat human employees. In his view, AI systems require access controls, cybersecurity protections, governance structures and oversight similar to those already used for workforce management. Agencies need to ensure AI agents can access the correct information while also preventing access to data they should not see. He says leaders must think carefully about how AI interacts with sensitive information, policies and operational processes.

The challenge, according to Cornelius, is no longer simply deploying AI technology. Instead, agencies must determine how humans and AI systems work together to achieve mission outcomes. He says employees need to understand how AI models function, how to improve them and how to provide effective guidance and feedback. In many ways, he explains, training an AI model resembles training an entry-level employee. Without clear instructions and continuous feedback, performance suffers.

Cornelius believes this shift is fundamentally changing how the federal workforce is defined. A few years ago, agencies often treated artificial intelligence as a specialized technical field staffed by a small group of dedicated AI experts. Today, he says, AI competency is becoming an expected skill across nearly every position description entering the federal hiring process. Agencies increasingly recognize that AI literacy is no longer optional.

He compares the transition to the governmentwide evolution of cybersecurity over the past decade. Initially, cybersecurity responsibilities belonged to a relatively small technical community. Over time, however, agencies realized every employee represented part of the cyber risk landscape, making cyber awareness a workforce-wide responsibility. Cornelius says artificial intelligence is following a similar trajectory. Instead of creating isolated AI teams, agencies now understand that employees throughout government need at least some level of AI competence.

This transformation is also driving changes in federal hiring and workforce management strategies. Cornelius says agencies are increasingly focused on skills-based hiring and skills-based workforce management rather than relying solely on traditional job classifications or lengthy resume reviews. Leaders want to understand what employees are actually capable of doing and how those capabilities align with mission needs.

He notes that many agencies still struggle to access reliable workforce data. During government workforce reviews and organizational changes earlier in 2025, Cornelius says some agencies could not easily identify employee skills, remote work arrangements or workforce composition because the information remained scattered across spreadsheets and outdated systems. That lack of visibility makes it difficult for agencies to manage workforce modernization effectively.

According to Cornelius, modern workforce management requires agencies to identify skills gaps, determine whether those gaps should be addressed through human talent or AI systems and ensure both humans and AI agents have the tools and access they need to succeed. The ultimate goal is mission performance, regardless of whether the work is completed by a person or an autonomous system.

At the center of Cornelius’s message is the idea that artificial intelligence is no longer a future capability waiting to arrive. Federal agencies are already integrating AI deeply into operational workflows, and workforce expectations are changing alongside the technology. As AI adoption accelerates, agencies must rethink hiring, training, governance and workforce strategy to ensure employees and AI systems can operate together effectively in support of government missions.