Technical note | Moving Forward with Computational Red Teaming
The term Computational Red Teaming has recently arisen within the literature to describe the application of new and innovative analytic techniques, tools and methodologies in support of Red Teaming Activities. This report explores Computational Red Teaming as a concept, which is itself undergoing transformation and growth within the practicing community. It describes just what Computational Red Teaming is, how it is applied, and its benefit over traditional Red Teaming practices and techniques. A framework of three key activities: Information Management; Conduct and Execution; and Scrutiny and Analysis; is then developed and decomposed into constituent functions for analysis.
The term Computational Red Teaming has recently arisen within the literature to describe the application of new and innovative analytic techniques, tools and methodologies in support of Red Teaming activities. The approach introduces novelty to Red Teaming, which is yet to be exploited; proposing to reduce risk and increase opportunities through computation.This report presents Computational Red Teaming as a concept, which is itself undergoing transformation and growth within the practicing community. Perspectives from the literature are presented to explore just what Computational Red Teaming is, how it is applied, and the perceptions of benefit over traditional Red Teaming practices and techniques.
Having explained the concept, and after reviewed the range of competing perspectives within the field; this report defines the term formally and develops a framework to categorise the field of Computational Red Teaming by the three key functions of: Information Management; Conduct and Execution; and Scrutiny and Analysis.
A full taxonomy is developed by means of functional decomposition1 . This taxonomy can be used to rationalise the application of Computational Red Teaming techniques within a program of work, task or an experimental campaign; by mapping or cross referencing the framework taxonomy elements against analogous manual tasks conducted within the program. This mapping can also be employed to identify where the functions of Computational Red Teaming can be applied, to add value or support program outcomes.The following recommendation is advised, for forward work planning within the Joint Operations Division's Computational Red Teaming Task.
The JOD Executive should consider initiating a scoping study of one targeted task within the division. This study would apply the taxonomy developed in this report, in order to identify where Computational Red Teaming tools and techniques could add-value (or otherwise contribute) to the task.
This research has been conducted as part of the JOD program of work into the study and development of Computational Red Teaming, following earlier work presented by Gowlett (2011). Gowlett called for the development of a divisional concept demonstrator for Computational Red Teaming. He argued that a niche capability might be developed in a specific targeted area. Should that effort prove successful, a wider program of research could be initiated.
Gowlett’s recommendations are still relevant today and development of this proposed prototype can be informed by our functional taxonomy. The greatest benefit is derived if a formal design methodology, such as Systems Engineering, is adopted because of the synergies between the structured approaches. The functional taxonomy is then employed as a cross reference against the design methodology. However, this is only meaningful if the design also employs functional decomposition. Systems Engineering is then a good candidate methodology for the design.
We reinforce the recommendation of Gowlett.
The JOD Executive should consider supporting the development of an executable prototype model for Computational Red Teaming. This prototype should be narrowly focused, as a concept demonstrator,and designed formally through adoption of Software or Systems Engineering practices. This approach will ensure best practice is followed in design and that minimal resources are consumed.
1 Functional elements are decomposed into their constituent components (Figures 1, 2, and 3 on pp. 22-24). The framework itself is also indexed to the fundamental sciences which underpin the functional elements and a set of critical enablers (Table 2 on p. 16).