Army adds Striveworks to its Next Generation Command and Control stack
The U.S. Army says software firm Striveworks is building the baseline artificial intelligence layer for the service’s Next Generation Command and Control architecture, or NGC2. The move adds a new piece to a larger modernization effort intended to unify battlefield communications and accelerate decision-making, especially as the Army prepares to expand the system from division-level testing to corps-level operations.
Army officials described the development at the TechNet Augusta conference, where they said the need for AI support will grow as NGC2 scales. At the corps level, the architecture is expected to process significantly more data than earlier prototypes handled, making the problem less about simple connectivity and more about how to turn large volumes of inputs into useful operational decisions.
Joseph Welch, the Army’s portfolio acquisition executive for command and control and counter-command and control, told Breaking Defense that the amount of software and data at corps scale will be greater and that AI tools will become more important in supporting that environment. That framing places Striveworks in a specific role inside the Army’s modernization plan: not replacing commanders or staff, but compressing the time needed to convert information into action.
From data foundation to decision layer
The Striveworks announcement follows an earlier June decision to have Anduril establish the baseline data layer for NGC2. Army officials have described that data layer as the foundation of the architecture because it creates a common baseline on which other software and services can be built and integrated. In practical terms, the data layer is about getting information into a shared structure rather than leaving it trapped in separate systems.
The AI layer sits on top of that foundation. If the data layer is meant to unify information flows, the AI layer is meant to help interpret them. That distinction matters because modern military networks do not suffer only from scarcity of data; they suffer from abundance. Sensors, drones and operational feeds can deliver more information than human staffs can manually process at useful speed.
Striveworks chief executive Jim Rebesco told the outlet that the company’s models will be able to compress hours of staff work into seconds. He also said today’s battlefields are saturated with real-time data and argued that the AI layer will help the Army match the best model to the right data while maintaining human oversight and governance. That last condition is central to how the Army is publicly describing the effort. The service is not presenting this as fully autonomous command software, but as an assisted decision framework with model management and oversight built in.
What Striveworks is expected to provide
According to Welch, Striveworks will be responsible for key functions that make an AI layer operational rather than theoretical. Those include model registry, validation, deployment and monitoring. Together, those capabilities form the backbone of an AI operations environment. A registry determines what models are available. Validation checks whether they perform as intended. Deployment makes them usable in operational settings. Monitoring tracks how they behave over time.
Those functions may sound administrative, but they are increasingly decisive in whether AI systems can be trusted in consequential settings. An army does not just need a model that works once in a lab. It needs a way to govern many models, track their performance, update them and ensure that specific algorithms are used for appropriate tasks. The Army’s description suggests it wants NGC2 to become a repository where soldiers and others can develop algorithms for specific needs and then make them available within a managed environment.
That is a different ambition from simply buying a single AI application. It points toward an architecture in which the Army can host multiple models and align them with specific missions and data streams. If realized, that would make NGC2 less a single software package and more an ecosystem for operational analytics and decision support.
Why corps-level scale changes the problem
NGC2 prototypes have already been tested at the division level with the 4th and 25th Infantry Divisions. The next step is larger. The Army plans in fiscal 2027 to field and scale the architecture within I Corps, headquartered at Joint Base Lewis-McChord in Washington, with oversight of Army units in the Asia-Pacific region.
That planned geography is notable. Pacific operations involve long distances, dispersed forces and potentially large volumes of sensor and communications data across maritime and land domains. Scaling command-and-control software in that environment is not merely a matter of adding more users. It means managing more data sources, more latency challenges and more operational complexity across a broader theater.
The Army’s emphasis on corps-level deployment also signals that NGC2 is moving from experimentation toward institutional use. Prototypes at division level can demonstrate concepts and identify bottlenecks. A corps-level rollout tests whether the architecture can function as a practical command backbone for larger formations. In that setting, data management and AI support become less optional enhancements and more structural requirements.
The broader meaning of the award
The Striveworks role shows how the Army is dividing NGC2 into interoperable layers rather than treating modernization as a single monolithic procurement. Anduril’s data layer and Striveworks’ AI layer suggest a stack approach: one company helps establish the data foundation, another helps operationalize model-driven analysis on top of it. That could make the architecture more modular, but it also raises the bar for integration and governance.
The announcement also reflects a wider defense trend. Military organizations increasingly describe AI not as a standalone capability, but as a function embedded within data pipelines, command systems and operational workflows. In that model, the main value comes from reducing staff burden, accelerating synthesis and helping users navigate information overload. Rebesco’s comment about compressing hours of work into seconds captures that logic directly.
Whether the Army can translate that promise into reliable field performance remains an open question. The supplied source text does not offer results from operational use at corps scale, only the Army’s plan and the company’s assigned responsibilities. But the direction is clear. NGC2 is being built as a layered digital architecture, and AI is no longer peripheral to that effort. It is becoming one of the service’s core mechanisms for handling the volume and speed of modern battlefield data.
If the fiscal 2027 I Corps rollout proceeds as planned, the Striveworks decision may be remembered less as a standalone contract announcement than as a marker of how the Army wants future command systems to function: data-rich, model-enabled and governed closely enough that commanders can use AI support without surrendering human control.
This article is based on reporting by Breaking Defense. Read the original article.
Originally published on breakingdefense.com
