In a significant move to enhance its operational efficiency, defense officials have recently utilized the Pentagon’s generative artificial intelligence platform to produce an impressive 100,000 agents. This initiative is part of a larger strategy by department leaders to accelerate the adoption of AI technologies across the Department of Defense (DOD).
The Pentagon first launched its GenAI.mil platform in December, aiming to equip millions of its personnel with modern commercial tools. Defense Secretary Pete Hegseth and CTO Emil Michael have both strongly advocated for this capability, encouraging its widespread integration within the DOD. Michael has emphasized the importance of applied AI, placing it high on his list of crucial technology domains.
Initial feedback during the rollout of GenAI.mil was mixed, but the platform has since gained acceptance across various military branches, including the Army, Navy, Marines, Air Force, and Space Force, as their preferred enterprise solution. Adoption has progressed positively, indicating a shift in mindset towards generative AI tools.
The first tools made available through the platform included Google’s Gemini products, with plans already announced to incorporate OpenAI’s ChatGPT and xAI’s Grok into the system in the near future.
Currently, over 1.2 million distinct users are engaged on the platform, as highlighted by Jacob Glassman, the deputy assistant secretary of defense for science and technology foundations in the R&E directorate, during a recent address at the Box Federal Summit organized by FedScoop.
Pentagon leadership has championed this technology as a transformative tool that can significantly enhance productivity and free up valuable time for personnel.
“We are in a fascinating time because, despite a workforce constraint due to the Deferred Resignation Program (DRP), we have laid the groundwork for AI with GenAI.mil,” Glassman articulated. He indicated that the Chief Digital and AI Office (CDAO) under the R&E undersecretariat has begun to release agents developed using this technology.
“In just two weeks, we successfully built 100,000 agents,” Glassman stated, emphasizing the swift progress in implementation.
The development of these tools comes at a critical moment given the ongoing ceasefire related to the U.S. conflict with Iran, known as Operation Epic Fury, which has increased demands on the DOD’s resources. “We are navigating a highly pressurized environment, as the conflict in Iran impacts our department almost daily,” he explained.
“With agentic AI at our disposal, we are witnessing improvements in efficiency and agility,” Glassman added. “This technology is empowering our workforce, allowing them to innovate under pressure.” He recounted a specific instance where a team faced a shortage of personnel while preparing a report for Congress. “I advised them to utilize GenAI.mil, and they returned with not just the report, but the quality was the best they had seen in five years,” he said.
During the conference, Glassman mentioned that department leaders are actively promoting the use of GenAI.mil, calling it a vital asset for the DOD.
The platform’s efficiency can alleviate mundane tasks, thereby allowing service members to focus on essential training and responsibilities. Glassman pointed out that he personally uses GenAI.mil “every single day,” and has found it consistently beneficial. “I have yet to have an instance where I regretted using it,” he noted, referring to his experience over the past five months.
He also commented on the importance of demonstrating AI integration to inspire broader adoption within the DOD, particularly among personnel and senior leaders in the R&E directorate.
That said, not all DOD employees may find the platform equally useful. “Given the diversity of our workforce, not everyone works in an office environment conducive to utilizing such tools. We have many mechanics and service personnel focused on hands-on tasks,” he admitted. However, Glassman expressed that he is more interested in the overall efficiency gains rather than the specific user statistics.
He acknowledged that precise metrics to measure the benefits of these technologies might be elusive, likening it to assessing the efficiencies achieved by transitioning from typewriters to personal computers.
