While the Department of Defense (DoD) has used artificial intelligence (AI) for decades, recent advancements create a new wave of opportunity and urgency. Interest has continued to grow, with the current administration prioritizing AI investments and aiming for expanded funding. Between fiscal years 2022 and 2023, AI-related contract obligations increased from $190 million to $557 million, signaling a broader focus on scaling AI in support of national security.
With growing investment and mission impact, both government and industry have an opportunity to shape how AI supports national security at every level. I recently joined leaders from the Defense Innovation Unit (DIU) and the Army for the GovCIO panel AI for National Security Applications, where we discussed how collaboration is bringing AI capabilities to the edge while safeguarding national interests.
Here are some of the major takeaways from our discussion:
Building strong foundations
Every impactful AI effort starts by getting the fundamentals right. As Mark Kitz from the U.S. Army’s Program Executive Office Command, Control, Communications, and Network (PEO C3N) explained, success starts with mastering the basics of data collection, curation, and accessibility. Before any AI solution can make a difference, organizations need to know where their data lives, how it moves, and what questions they are trying to answer. From there, it becomes possible to ask better questions and create tools that lead to better decisions.
Agencies can prepare by integrating data science, analytics, AI, and machine learning into their environments and partnering with experienced integrators focused on building AI-ready infrastructures that serve real-world missions and real-world outcomes. Key to this approach is leveraging proven enterprise platforms to create solutions that shape better, faster, and more informed decision-making.
Designing environments for the warfighter
AI must fit where it will be used and needed most. For example, warfighters rely on lightweight, adaptable models that can operate with limited bandwidth, alongside older systems, and in tough conditions. Jamie Fitzgibbon, AI/ML Program Manager for DIU, pointed out that new AI capabilities need to work within these existing environments, not in perfect lab conditions.
Collaboration is key from the industry side. Industry is most effective when we work closely with defense partners to replicate realistic conditions like restricted compute power and unique security requirements. This approach helps ensure that AI solutions and development services address the realities of field operations, not theoretical use cases. Advanced automation, including robotic process automation (RPA) technologies, can help transform systems to be resilient, adaptable, and meet the demands of the environment, whether operating from a command center or deployed at the tactical edge.
Preparing the workforce for an AI-powered future
Deploying AI in services of national security missions is a human-centered effort. Success requires a change in how organizations think about workflows, decision-making, and skills development; it depends on ensuring everyone understands how these tools fit into their roles, decisions, and program outcomes.
For instance, not everyone needs the same level of technical knowledge. Leaders need to know what questions to ask, and users need to understand how AI tools fit into their missions. Legal teams, Human Resources teams, and acquisition specialists play important roles in integrating AI into daily work. Actual progress comes from preparing the entire workforce, not a select few.
Moving toward multi-agent systems
Emerging Agentic AI systems, in which multiple AI agents coordinate actions and functions across workflows, processes, and applications, are opening new possibilities for defense operations. These new models allow AI agents to make decisions, collaborate, and orchestrate insights-based actions with some autonomy. They can bring new levels of speed, adaptability, and resilience to complex missions.
Industry can serve government agencies by working alongside them to develop the frameworks, governance, and safeguards needed to responsibly bring multi-agent systems into real-world operations. Success depends on thoughtful design, responsible AI use, and understanding of mission priorities.
Discover more
AI for national security is happening now and will continue to reshape how missions are executed and supported. Success depends on collaboration across government, industry, and the workforce, with a shared vision of building mission-driven outcomes that are resilient and ready for tomorrow’s challenges.
Watch the full interview now to hear more insights, including perspectives on data sharing, infrastructure, human training, and new AI models.