Presented by Snowflake

On Edge to Enterprise: Inside the Department’s Data Revolution, host George Jackson explores how defense and industry leaders are responding to the department’s push toward an AI-first warfighting force. Presented by Snowflake, the program looks at the data foundations, governance models, security requirements and edge-computing realities that will determine whether artificial intelligence can move from promising pilots to trusted mission capability across the Navy, Marine Corps, National Guard Bureau and the broader joint force. 

Building the Data Foundation for AI-Driven Mission Advantage

In the opening segment of Edge to Enterprise: Inside the Department’s Data Revolution, host George Jackson frames the conversation around the Pentagon’s accelerating push toward artificial intelligence and the enterprise data strategy needed to support it. Michael Frank, Deputy Chief Technology Officer at the Department of the Navy’s Office of the Chief Information Officer (DoN CIO), explains that the Department of the Navy’s AI adoption efforts are focused heavily on the rollout of GenAI.mil, an unclassified and broadly accessible capability designed to help users begin integrating artificial intelligence into daily workflows. For Frank, the value of this early implementation is not only operational; it is cultural. By giving personnel a practical, low-risk way to interact with AI, the department can begin building trust in the technology, helping users understand how prompts produce outputs, where the systems are reliable, and where human judgment remains essential.

Jennifer Chronis, Vice President of U.S. Public Sector at Snowflake, expands on that point by emphasizing that there is no meaningful AI strategy without a strong data strategy. She explains that organizations that prioritize the data layer first are the ones best positioned to see real value from
AI. That means creating data environments that are unified, governed, secure, shareable and discoverable. Chronis also stresses the importance of open data formats, noting that government organizations should not be locked into a single vendor or forced to “rent” access to their own data. When mission needs, contracts or requirements change, the government must retain control and portability of its data.

The conversation then moves from technology into culture. Frank notes that AI adoption is not simply a technical modernization effort; it requires a structural shift in how people think about their role in decision-making, the kill chain and the OODA loop. He argues that the human relationship with technology is changing, and that trust must be earned through practical experience, especially in unclassified environments where the risks are lower but the lessons are valuable.

The segment also highlights the increasing importance of data sharing across services, commands, allies and partners. Chronis explains that modern technology can reduce the risks and delays historically associated with moving and copying data by enabling governed, real-time access without unnecessary duplication. Frank adds that joint and coalition interoperability is no longer optional. In a large-scale conventional fight, no single service — and not even the United States alone — can operate effectively without the ability to share data securely across organizational and classification boundaries.

Finally, Frank connects the data conversation to rapid commercial technology adoption. As the Department of the Navy looks to commercial-first approaches, unmanned systems and emerging capabilities, he cautions that hardware alone will not deliver mission advantage. Without the communications infrastructure and data backbone to support those systems, the department risks investing in tools that cannot perform when needed. Together, Frank and Chronis make the case that AI adoption begins with trusted, open, governed and mission-ready data.

Key Takeaways

• AI adoption depends on trust as much as technology. Michael Frank explains that GenAI.mil gives Department of the Navy users a practical way to build familiarity with AI in everyday workflows before applying it to more sensitive or mission-critical use cases.
• A strong data strategy is the foundation for AI outcomes. Jennifer Chronis emphasizes that unified, governed, secure, shareable and open data is essential for defense organizations that want to move quickly and avoid vendor lock-in.
• Joint and coalition data sharing is now table stakes. The segment makes clear that future operations will require data to move securely across services, commands, allies and partners, especially in complex and contested environments.

From Enterprise to the Edge: Data Sharing, Security and Resilience Across the Mission

The second segment of Edge to Enterprise: Inside the Department’s Data Revolution shifts the discussion from enterprise data foundations to the operational realities of sharing information across the joint force and out to the tactical edge. Host George Jackson is joined by Jeffery Hurley, Acting Director of IC4 and Deputy CIO for the U.S. Marine Corps; Kenneth McNeill, CIO/J6 and Director at the National Guard Bureau; and Adam Edelman, Senior Manager for Solution Engineering for the Department of War at Snowflake. Together, they examine how data-centric operations are being put into practice in environments where speed, security, resilience and mission partner access all matter at once.

Hurley begins by describing the Marine Corps’ focus on getting the right data to the right place at the right time, especially for Marines operating at the tactical edge. He explains that the Marine Corps has established a service data office to help break down silos and improve access to trusted information. But he also points to a persistent challenge: data labeling and source-of-truth issues. When multiple systems produce multiple updates to the same information, even AI tools can struggle to determine what is accurate. For Hurley, solving that problem is central to making data usable in the field.

McNeill then discusses Project Homeland, which he describes as the National Guard Bureau’s contribution to CJADC2 in the homeland mission space. Unlike overseas warfighting scenarios that may focus primarily on federal and military partners, the National Guard must share information with state and territorial governments, local first responders and other civilian organizations during events such as hurricanes, storms, floods and major public gatherings. McNeill explains that Project Homeland is designed to help the National Guard communicate across those boundaries while still maintaining connectivity with federal partners. He notes that major events, including the FIFA tournament, provide an important proving ground for multi-state, multi-city information sharing.

Edelman connects these operational examples to the broader technology shift around data sharing. Drawing on his previous experience as an analyst, he says data sharing once felt risky because data that left an organization’s boundary was effectively gone. Today, however, technology allows organizations to share access to data without copying it. That approach gives analysts and mission partners the ability to collaborate while preserving governance, control and auditability.

The segment also addresses the infrastructure and security demands of AI at the edge. Hurley explains that AI requires massive compute, storage and data resources, but Marines in distributed environments cannot carry a data center with them. The challenge is to deliver useful AI capability in smaller, more ruggedized forms that can operate even when disconnected from satellites or the internet. McNeill reinforces that security remains the top priority, particularly for the National Guard’s broad attack surface across 54 states, territories and the District of Columbia. He points to zero trust, workforce training and recurring exercises as essential measures for building confidence and readiness.

The discussion closes with a focus on resilience. Edelman argues that cloud alone does not solve resilience because cloud infrastructure still depends on physical locations, vendors, models and tools. Hurley agrees, explaining that the department has learned the limits of relying on a single cloud provider or architecture. True resilience requires a mix of home networks, cloud capabilities, multiple vendors and portable data that the government continues to own and control as missions change.

Key Takeaways

• Getting data to the edge requires more than connectivity. Jeffery Hurley explains that the Marine Corps must solve data labeling, source-of-truth, compute, storage and disconnected-operations challenges to make AI useful for Marines in the field.
• Project Homeland shows why mission partner data sharing matters. Kenneth McNeill describes how the National Guard Bureau must exchange information with federal, state, territorial, local and first-responder partners during domestic missions and major events.
• Resilience now includes data portability, vendor flexibility and workforce readiness. Adam Edelman and the government panelists make clear that secure, governed access to data — without copying or locking it into one platform — is essential for mission success.