network topologies, simulate cy - berattacks, and assess defensive re - sponses in real time. This approach represented a major leap forward from traditional static cyber ranges. Rather than relying on prescripted attacks, CASTLE’s adaptive agents continuously evolved tactics, forcing defenders—including National Guard cyber teams and DEVCOM engineers— to respond to unpredictable and so - phisticated adversarial behavior. The integration of CASTLE into Cyber Fortress exemplified the prin - ciple of “testing early and often” in the innovation cycle. By embedding DARPA’s experimental technology di - rectly into a live operational exercise, the program accelerated feedback loops between researchers, indus - try technologists, and Warfighters. Cyber operators provided real-world insights into system usability, data fidelity, and operational tempo— feedback that DARPA engineers could immediately incorporate into iterative development. This collaboration bridged the traditional divide between research laboratories and operational units. Instead of waiting for technologies
The defense community increasingly recognizes that depending on a centralized grid creates a vulnerability: Outages, cascading attacks, or deliberate electromagnetic interference can compromise or cripple operations.
to mature through years of isolated R&D, Cyber Fortress allowed CASTLE’s AI agents to evolve within a realistic, contested environment. The exer - cise validated not only the technical performance of CASTLE’s algorithms but also their operational relevance in defending critical infrastructure and mission networks. Lessons From Cyber Fortress The Cyber Fortress exercise was more than a technical demonstra - tion—it was a systems-level experi - ment in how the DoW, state govern - ments, industry, and academia can collectively accelerate innovation in the face of evolving threats. The exer - cise also underscored a critical truth:
Technology alone doesn’t guarantee readiness—cooperation and dynamic integration do. As defense operations expand across domains—cyber, space, elec - tromagnetic spectrum, and critical in - frastructure—alliances between pub - lic and private sectors will become the primary engine for innovation. The insights gained from Cyber For - tress reveal what works, what needs work, what must change, and how the defense community can evolve to meet the demands of tomorrow. Test early and with Warfighters. Integrating DARPA’s CASTLE program and other early-stage R&D efforts into the exercise shortened the innovation feedback cycle. By testing emerging capabilities with operational users present, Cyber Fortress replaced the traditional “research-then-field” model with one that presents imme - diate challenges and validates tech - nologies in real time. Operate outside stovepipes. Par - ticipants repeatedly noted that the most valuable breakthroughs weren’t only technological but cultural. Cross- agency networking, joint intelligence sharing, and trust-building events— such as the Distinguished Visitor days, tabletop exercises, and force-on-force cyber play—created the organic con - nective tissue that technology alone cannot provide. Measure and scale. For innova - tions such as microgrids synchroniz - ing and managing a portfolio of en - ergy assets and perovskite modules to gain sustained investment, per - formance metrics must be rigorously captured and analyzed. Key indica -
Participants get down to business online in the Virginia National Guard 2025 Cyber Fortress exercise. Source: Author
MAY – JUNE 2026 | DEFENSE ACQUISITION MAGAZINE 35
Made with FlippingBook - Online Brochure Maker