NGC2’s AI command shift moves into procurement

The Army’s shift to Next Generation Command and Control began as a break with a badly fragmented baseline. Before 2024, the service relied on “a federation of service- and echelon-specific command-and-control systems and data repositories that did not automatically share data or provide a common operational interface,” but between 2023 and 2024 Army Futures Command directed an experimentation effort to replace those fragmented, platform-specific tools with “a software-centered architecture that fused data, AI-enabled applications, and resilient communications” for disrupted or low-bandwidth conditions.

What changed in 2023–2024 was not just an integration layer but a redesign of the command-and-control model itself. The effort “displaced this fragmented legacy approach with a clean-slate, modular, data-centered architecture,” and the Army showed that transition in practice when “this effort was first publicly demonstrated at Project Convergence Capstone 4 in March 2024,” underscoring that NGC2 had become a software-centered architecture for preserving situational awareness even when connectivity was contested rather than a concept confined to planning documents.

The Army is consolidating AI-driven command and control capabilities by building a software layer above disconnected systems so they can function as one decision environment. In the September interview, that AI abstraction layer was described as pulling information from different places, consolidating it, making decisions on it, and presenting it to users without requiring them to know which application is talking to which, while also letting the government modernize underlying applications behind the scenes without disrupting operators or breaking interoperability across vendors.

That architecture matters because the operational problem is less collecting data than turning fragmented, incomplete inputs into timely choices across complex environments. Anduril has framed the battlefield as a place with thousands and Thousands of systems generating more information than humans can coordinate, and the October interview on multi-agent AI offered a close analogue: complex systems require thousands of suppliers to build a factory, where if one specific part arrives 2 weeks later, it can cause hundreds of millions of dollars in damage, and a $100 million loss may be hidden unless someone decides it matters; similarly, one supplier with 20% of the goods missing versus 1% missing may still be worth choosing if it is 10x cheaper, but reliability can justify paying 10x the amount for a given use case.

The October awards look less like another experimentation cycle than the moment NGC2 became a funded procurement program with duration, scale, and repeatable budget lines. PR Newswire reported on Oct. 8, 2026 that the Army “has awarded Striveworks a five-year, $200 million contract” to extend AI capabilities across the Indo-Pacific for NGC2, a structure that by itself signals a break from one-off prototype buys and toward a sustained acquisition path built around recurring obligations over multiple years.

Just as important, the Striveworks award ties that money to operational scaling rather than trial activity: under the contract, the company will “scale its suite of Chariot software across the NGC2 footprint” and provide “common AI tools and services for NGC2, enabling the Army to deploy, monitor, and re-optimize AI models and agents” as conditions change. PR Newswire said the award also “falls under the company’s previously executed enterprise contract,” and “follows the service’s confirmation in August” that Striveworks would provide “baseline AI tools and services,” reinforcing that October’s surge was plugging NGC2 into an ongoing program of record.