Defense organizations don’t lack data. They generate billions of data points across sensors, aircraft, ships, satellites, logistics systems, intelligence platforms, cyber tools and personnel in the field. The harder problem is making the right information usable across the systems and organizations that need it, moving it without losing the context, security controls and provenance that make it trustworthy.
That challenge is central to Joint All-Domain Command and Control, or JADC2. The Defense Department’s framework is intended to improve how information is shared across military services, operational domains and mission partners, so commanders and warfighters can understand the operating environment and make decisions faster.
Artificial intelligence raises the stakes. As defense agencies incorporate AI into analysis and decision-making, the effectiveness of those capabilities will depend on the information they can access. Models working with incomplete, outdated or poorly governed data risk producing an equally incomplete picture of the mission; AI doesn’t fix underlying data limitations, it turns them into analytical limitations at machine speed.
For JADC2 and AI to advance together, data readiness has to become a mission requirement, not an afterthought.
AI has to work across a distributed data environment
Defense agencies are implementing JADC2 across technology environments built over decades. Mission data is spread across legacy and modern systems, data centers, public and private clouds, secure enclaves and tactical platforms. Some workloads also need to stay close to sensors and operational units, while security classifications and policy requirements constrain where sensitive information can be stored and processed. That combination makes moving every dataset into a centralized environment impractical, introducing latency and security concerns, in addition to consuming significant time and resources.
Leaving information isolated within individual systems creates a different problem. An April 2025 Government Accountability Office review described this as a “swivel chair” model, in which personnel receive information through one system, assess it manually and then transfer or communicate it through another, a process GAO found makes it harder for commanders to decide at the scale and tempo modern missions require.
When sensors, networks, software and AI can exchange and act on data more directly, time to insight can be cut significantly, creating great tactical edge capabilities for defense operations.
Interoperability has to extend to the data
Reaching data is only half the equation when it comes to successful JADC2 implementation; the systems requesting it also need to be able to use it once it arrives. JADC2’s interoperability challenge reaches across hundreds of critical systems developed for different services, missions and technical requirements. Combined Joint All-Domain Command and Control (CJADC2) compounds that challenge by extending the information-sharing environment across the intelligence community, international allies and coalition partners.
The same GAO review highlights the difficulty of connecting those systems across a contested battlespace amid service-specific requirements, stove-piped development and proprietary architectures. Defense Information Systems Agency officials have also pointed to a related wrinkle: Partners and allies maintain their own data repositories and sit at different stages of building out secure networks, complicating any push toward common standards.
Connectivity alone doesn’t guarantee usability. Two systems can technically exchange information and still interpret it differently if they rely on incompatible formats, schemas or metadata. For AI systems attempting to correlate information across those environments, those inconsistencies become a data-quality challenge as much as an interoperability one.
Consider a tsunami response in the Pacific involving multiple U.S. military services, local authorities and humanitarian organizations. Each participant brings different systems and information to the mission, but all need a sufficiently shared view of conditions to coordinate personnel and resources, the kind of scenario where AI-assisted analysis is only as good as the data it can reach.
A previous Federal News Network commentary on coalition readiness pointed to the Defense Department’s Modular Open Systems Approach and Sensor Open Systems Architecture as efforts to improve interoperability. Open architectures and shared standards can enable independently developed systems to exchange information, but that connectivity has to extend to the data itself. Common standards help systems exchange information more consistently, but the real objective goes beyond connecting networks: Data needs enough shared structure and context that people, applications and AI systems can understand and use it across organizational boundaries.
Governance determines what information reaches the mission
Making data accessible does not mean making all data available to everyone. Every additional organization brings its own systems, data, classifications, policies and authorities, and access requirements have to shift as participants move from one mission to another.
Bold Quest 22 demonstrated the scale of the access challenge. More than 1,000 participants from the U.S. military, allies and partners across 19 countries tested CJADC2 capabilities and coalition interoperability. The exercise advanced technical and procedural interoperability among participating forces, while giving teams an opportunity to identify shortcomings that could prevent systems and information from working together effectively before those capabilities reached the battlefield.
Role- and attribute-based controls can help ensure users receive information appropriate to their identity, mission and authorization. Governance also needs to preserve provenance, where information originated, how it moved between systems and what happened to it along the way.
That balance is harder to strike than it sounds. The GAO found that overly restrictive classification hinders command-and-control data sharing, including with allies and mission partners, and cited specific friction points. One government database considered critical to a trilateral security partnership couldn’t be shared more broadly once it carried that critical designator. And even where the U.S. has data-sharing agreements with two allies individually, it can’t combine their data in a coalition environment if those allies lack an agreement with each other. GAO also found that once information is tagged with a restrictive designator, changing it later is difficult even after officials agree the restriction no longer serves its purpose.
That challenge carries added weight once AI informs the analysis built on top of the data. Expanding access alone isn’t enough; commanders need confidence that what’s informing an AI-generated analysis is authoritative, properly governed and appropriate for their use.
Start with the data foundation
The pace of AI development is creating pressure to get new capabilities into the hands of warfighters quickly. But deploying a new model doesn’t make the underlying data environment ready for AI. Defense leaders also need to weigh whether the data environment underneath it can safely support the mission, especially when lives are on the line.
That means asking practical questions. Can warfighters find and trust the information relevant to a decision? Can services share data with coalition and mission partners while enforcing the appropriate controls? Can analytics and AI continue operating when connectivity is constrained? And can agencies trace where the information behind an AI-generated output originated?
For JADC2, that makes the data architecture more than an IT consideration, it is part of the mission infrastructure. As JADC2 and AI evolve together, ensuring trusted information can reach people and systems wherever decisions are being made will be fundamental to realizing the promise of both.
Jonathan Veal is group vice president of defense and intelligence at Cloudera.
Copyright © 2026 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.