Fighting with Data: Design Implications for AI-Enabled Mission-Command Systems : Army War College , August , 2026
Fighting with Data: Design Implications for AI-Enabled Mission-Command Systems : Army War College , August , 2026
From the report: "Fighting with Data argues artificial intelligence (AI)–enabled mission-command systems do not simply make existing military decision-making processes faster; they fundamentally redistribute cognition across a human-machine network. In legacy systems, commanders and staffs rely on hierarchical, sequential processes in which humans manually collect, correlate, interpret, and act on information. In AI-enabled systems, machines assume a much larger share of the cognitive labor associated with sorting data, detecting anomalies, correlating signals, and generating candidate explanations. As a result, the role of the human shifts from assembling the picture to interpreting the picture’s meaning, judging risks, and deciding how to act.
This shift in the human role creates challenges in two areas of AI-enabled mission command: trust and optimization. The first challenge concerns the limits of reliabilist approaches to machine output. In conventional settings, trust can be grounded in factors such as data curation, model validation and verification, robustness, calibration, and expert oversight. But these indicators are least helpful when AI output is future oriented—when systems are used to generate plans, predictions, or assessments of enemy intent whose truth cannot be known in advance and may remain contestable even after the fact. For example, a failed plan may still be the best one available under the circumstances, whereas a successful plan may rest on spurious correlations or opaque reasoning. Regarding optimization, most current approaches focus too narrowly on the relationship between human and machine, neglecting the larger distributed cognitive system through which machine output becomes knowledge and understanding. Thus, the real design challenge is not merely to optimize model performance. Instead, the challenge is to optimize the fit of the entire human-machine system to the operational environment."
Authors - Pfaff, C. AnthonyRelated Resources
