DIA Envisions AI-Powered 'Agent-to-Agents' Interactions for Military Operations

The Defense Intelligence Agency (DIA) is working towards a future where artificial intelligence tools support military operations through seamless interactions between digital agents. According to Maj. Gen. Robert Kinney, the DIA’s chief AI officer, this vision involves ‘agent-to-agents’ interactions that enable data analysis and decision-making across combatant commands.

Kinney shared his insights during a panel at the DODIIS Worldwide conference in Tampa, Florida, where he discussed the agency’s ongoing efforts to build an AI enterprise platform service. This 90-day ‘sprint’ aims to establish a comprehensive framework for integrating AI tools and data sources within the military intelligence community.

One key component of this effort is the model context protocol (MCP), which will provide a standardized way to access intel data from various external systems and data sources. MCP serves as the technical foundation that enables AI tools to connect with these external systems, facilitating more efficient data exchange and analysis.

Looking ahead to the next year, Kinney emphasized the importance of building digital agents that can interact with each other in a secure and compliant manner. This involves developing ‘agents-with-agents’ capabilities, which will allow for more sophisticated decision-making and task execution within military operations.

The potential benefits of this approach are significant, as it could enable real-time data sharing and analysis between different directorates and commands. For instance, an agent responsible for collection management might communicate with agents in the 3 Directorate (operations) or the 4 Directorate (contested logistics), facilitating more informed decision-making across the board.

However, officials must also consider the risks associated with these emerging technologies. Kinney noted that setting up guardrails to mitigate potential risks is crucial, particularly when it comes to irreversible mission areas such as fires and planning. In such cases, human oversight will be essential to ensure responsible use of AI capabilities.

The DIA is currently laying the digital foundation for this vision by establishing ‘pipes’ that support data access and agent development. Additionally, officials are exploring the ‘tradecraft piece’ of using agents interacting with each other, focusing on responsible parameters for operating within these systems.