Three Things I Learned Speaking at the AI for National Defense & Security Summit
I had the opportunity to participate in an AI for National Defense and Security event focused on how government and industry are operationalizing artificial intelligence across the defense and intelligence communities. The discussion brought together leaders from across the national security ecosystem to address a common reality: AI is no longer a future concept for the Department of Defense and the intelligence community. It is already being integrated into operational environments that directly impact mission execution, decision advantage, cybersecurity, logistics, intelligence analysis, and coalition operations.
What stood out most during the event was not simply the pace of AI adoption, but the growing realization that the next phase of AI implementation will depend heavily on trust, security, interoperability, and governance. The conversation has clearly evolved beyond experimentation. The challenge now is how to operationalize AI responsibly at scale inside highly contested and highly classified environments.
As I reflected on the discussions afterward, three major themes emerged that reinforced where the defense community is heading next.
1. Zero Trust Is Rapidly Becoming Foundational to AI Operations
One of the strongest themes throughout the event was that artificial intelligence cannot scale effectively inside the Department of Defense without a modernized security architecture built around Zero Trust principles.
For years, Zero Trust discussions largely focused on identity, endpoint security, and access management. That conversation is now expanding toward protecting AI models, training data, inference engines, and mission decision systems. Government leaders are increasingly recognizing that AI systems themselves must become part of the Zero Trust ecosystem.
The emerging challenge is no longer simply:
“Can we deploy AI?”
The challenge now, is:
“Can we securely operationalize AI across multiple classification levels, coalition environments, and mission systems without introducing unacceptable risk?”
That shift has major implications for the national security community. AI models require access to massive amounts of data, but not all data can be freely shared across domains, organizations, or allied partners. As AI adoption accelerates, the ability to securely move, validate, govern, and protect data becomes mission critical.
This is where technologies like Cross Domain Solutions (CDS), Insider Risk Management, continuous verification, and Data-Centric Security (DCS) become increasingly important. AI will only be as trustworthy as the security architecture surrounding it.
The Department of War Zero Trust Portfolio Management Office has already begun signaling publicly that the next evolution of Zero Trust will extend into operational technologies, AI-enabled environments, and data-centric mission systems. The direction is becoming increasingly clear: AI and Zero Trust are no longer separate conversations.
2. Data Readiness Is Still One of the Biggest Obstacles to AI Success
Another consistent theme from the event was that many organizations are still struggling with the foundational challenge of data readiness.
There is tremendous excitement around large language models, AI-enabled analytics, autonomous systems, and mission acceleration. However, several speakers emphasized that AI effectiveness ultimately depends on the quality, accessibility, classification, governance, and integrity of the underlying data.
In national security environments, this challenge becomes exponentially more difficult because the data often exists:
- across multiple security domains,
- inside disconnected legacy systems,
- within coalition environments,
- or under strict releasability restrictions.
Many organizations are realizing that simply deploying an AI model does not solve these problems. In some cases, AI exposes how fragmented and siloed existing data environments have become.
What became clear during the event is that the defense community is shifting toward a broader concept of “trusted data mobility.” Organizations are beginning to recognize that secure data movement, policy enforcement, metadata tagging, and access governance are foundational requirements for successful AI operations.
This is particularly important for coalition operations and Joint All-Domain Command and Control (CJADC2) environments, where decision advantage depends on moving the right information to the right operator at the right time without compromising mission security.
The future of AI in defense will not belong solely to the organizations with the biggest models. It will belong to the organizations that can securely operationalize trusted data at scale.
3. Government and Industry Partnerships Will Determine the Speed of AI Adoption
The third major takeaway was the importance of collaboration between government and industry.
No single organization can solve the AI challenge alone. The operational requirements are simply too broad and too complex. The conversations yesterday reinforced that the future national security architecture will depend on an ecosystem of government agencies, traditional defense contractors, next-generation defense technology companies, cloud providers, AI innovators, and cybersecurity leaders working together in a much more integrated way.
What I found encouraging was how much the conversation has matured around interoperability and partnership-driven innovation. There is growing recognition that AI adoption must occur inside secure, mission-aligned ecosystems rather than isolated point solutions.
Partnerships between companies focused on AI, Zero Trust, cyber defense, insider risk, secure mobility, and operational mission systems will become increasingly important as agencies move from pilot programs into enterprise-scale deployments.
The reality is that AI adoption inside national security environments introduces new attack surfaces, new insider risk concerns, new governance requirements, and new operational dependencies. Successfully navigating that environment requires deep collaboration between organizations that each solve different parts of the mission challenge.
The organizations that move fastest over the next several years will likely be the ones that build strong technology alliances and operational ecosystems early.
Final Thoughts
The conversations at yesterday’s event reinforced something that is becoming increasingly apparent across the defense and intelligence communities:
Artificial intelligence is no longer a standalone technology initiative. It has become part of the operational fabric of national security.
That said, operationalizing AI responsibly requires more than advanced models. It requires trusted architectures, secure data mobility, Zero Trust enforcement, insider risk protections, coalition interoperability, and continuous governance.
The national security community is entering a new phase where the conversation is shifting from:
“What can AI do?”
to:
“How do we securely deploy AI at mission scale in contested environments?”
That is a far more important conversation, and one that will shape the future of defense modernization for years to come.