Ryan Nguyen is the artificial intelligence capability lead at Arcfield.
In 2025, federal missions underwent rapid evolution, with an increase in data volume, decision-making speed, and engineering complexity. To manage these challenges, teams needed technology that reduced cognitive load rather than adding to it. Artificial Intelligence (AI) demonstrated potential, but its effectiveness depended on targeted use. This guiding principle shaped Arcfield’s mission support throughout the year, driving progress across the programs we served.
Arcfield focused on assisting engineering teams in extracting, refining, and navigating rapidly changing requirements and system models. Leveraging our deep expertise, we rethought systems engineering workflows through a close partnership between our model-based systems engineering (MBSE) subject matter experts and AI innovators. Through a sizeable internal research and development investment, Intelligent MBSE was created. By harnessing the power of AI throughout the entire MBSE process, Intelligent MBSE significantly accelerates engineering processes by identifying issues early, expediting document analysis with minimal disruption, navigating large system model repositories, and maintaining alignment with mission intent amid changing conditions. This approach was built on the premise that reimagining workflows, rather than merely automating individual steps, is key to unlocking AI’s efficiency potential.
I would argue that information integrity is even more critical than ever. With synthetic content, blended sensor data, and model-assisted analysis becoming more prevalent, missions require robust approaches to provenance and validation. Establishing a solid framework of truth and clear connections between assumptions and outcomes enables engineering teams to make decisions with greater confidence, and we expect these practices to expand across mission workflows.
A key lesson from 2025 is that while many organizations pursue AI primarily to mimic or replace human behavior, enduring advantage comes from deeper integration that fundamentally reshapes workflows. This distinction is especially consequential in the national security domain. Many adversaries are rapidly deploying AI across military and intelligence systems, often focusing on automating existing processes rather than rearchitecting them for true machine-native integration.
This tendency creates a strategic opportunity. When AI systems are designed to imitate human decision-making within legacy workflows, they tend to exhibit predictable behaviors and shared structural weaknesses. By understanding the technological foundations and common implementation patterns of such systems, we can anticipate where they are most vulnerable—not just technically, but operationally.
That understanding enables the development of offensive counter-AI approaches aimed at exploiting these weaknesses. By analyzing how adversarial AI systems are trained, integrated, and operationalized, we can identify recurring patterns, manipulate inputs, disrupt feedback loops, and degrade decision-making processes at scale. The objective is not merely to build superior AI, but to design systems and tactics that specifically undermine the effectiveness of poorly integrated or hastily deployed AI.
Although this discipline is still emerging, it represents a growing opportunity to turn the inherent limitations of adversarial AI into tactical and strategic advantage. As operational environments increasingly shift toward machine-to-machine interaction, success will depend less on raw AI capability and more on a nuanced understanding of where AI fails—and how those failures propagate.
This offensive counter-AI strategy marks an important evolution in modern warfare. Rather than simply matching adversaries system for system, it emphasizes negating their investments by targeting the predictable flaws of human-imitative AI implementations. Done correctly, this approach allows the United States to degrade adversarial AI effectiveness rapidly, while preserving the resilience and mission advantage of its own more deliberately designed, deeply integrated systems.
These trends are influencing how agencies approach acquisition. As AI becomes more integral to operations, customers are asking more precise questions about transparency, resilience and long-term maturity. This year has demonstrated that mission success hinges on technology that reduces friction, supports personnel, and clarifies complex tasks. In 2026, we aim to sustain this momentum with thoughtful AI integration that strengthens decision-making and enhances mission operation speed and resilience.