Salesforce, headquartered in San Francisco, has announced that its autonomous AI agents have received approval from the Defense Department and the Army’s Human Resources Command for use in handling sensitive military operations.
On Wednesday, the company disclosed that its Missionforce National Security platform secured Impact Level 5 (IL5) authorization, which signifies its capability to manage the most sensitive unclassified military missions.
Kendall Collins, CEO of Missionforce and Government Cloud at Salesforce, emphasized that the platform is part of a broader initiative to modernize government information technology, asserting that they are establishing a reliable AI framework for national security.
The Missionforce platform is a government-tailored version of the commercial Agentforce platform, featuring a network of AI agents designed to enhance analysis and decision-making processes.
This technology rollout follows Salesforce's substantial $5.6 billion contract awarded in January to develop Missionforce. In July, the company announced that the Air Force had implemented Missionforce within the 441st Vehicle Support Chain Operations Squadron to streamline management of its $13.5 billion fleet, which comprises over 84,000 vehicles spread across nearly 389 locations.
Salesforce characterized this new approval as a significant achievement, marking a progression for national security and defense operations from basic predictive AI and chatbots to full-scale autonomous AI execution. The AI agents will have access to a mix of controlled unclassified information alongside data from the unclassified National Security System.
Although Missionforce is specifically designed for military applications, it is built on the foundation of Salesforce’s commercial Agentforce 360 platform, which was launched in late 2025.
Ryan Von, a writer for Salesforce Ben, noted that Agentforce represents a major advancement in the company’s AI capabilities, referring to the latest 360 variant as “the most powerful version of the tool yet.” He highlighted its ability to provide accurate, contextually aware responses in both text and voice formats.
In preliminary evaluations of the commercial platform, Von observed encouraging results, although he mentioned that while the AI excelled with straightforward, keyword-driven queries, it encountered challenges with more nuanced inquiries, such as personalized recommendations.
For the Army’s Human Resources Command, Salesforce indicated that the AI agents would deliver continuous support, handle routine inquiries, summarize case histories, and provide relevant policy and career information sourced from Army-approved resources.
The Human Resources Command processes over 1,500 personnel cases daily, and Salesforce estimates that the deployment of AI agents could save approximately $6 million each year while managing over 55 million conversations monthly.
Bill Pessin, Senior Vice President of Missionforce National Security at Salesforce, noted that the platform would enable human personnel to concentrate more on intricate, career-impacting decisions that directly influence the readiness and lives of soldiers, alleviating them from mundane tasks that can be automated by AI.
Daniel Terrill, a contributor to Military Times, has covered military topics, the gun industry, and outdoor activities for nearly 20 years. In addition to his writing career, he has experience as a Marine, police officer, and even worked as a substitute teacher, which he considers one of the most challenging roles he has taken on.
