Active–Reactive Deception
An adaptive system reveals more than a single decision. Its timing, adjustments, and response to changing conditions can carry useful evidence about how it operates. This research explores controlled interactions as a way to make that evidence observable.
The question
What can a defender learn by changing the conditions of an interaction, rather than only watching it unfold?
Observation becomes an experiment
The work treats a response as part of the measurement. A controlled change creates a comparison: what happened before, what changed, and how the system reacted. The goal is to learn from behavior while keeping the interaction within a defined operating boundary.
The response needs context
A reaction can have several explanations. Timing, the task being attempted, the surrounding conditions, and ordinary variability all matter. Useful evidence must survive those competing explanations.
Why it matters
This connects behavioral detection to decision-making. The next observation can be chosen for the information it is likely to provide, with the cost to legitimate activity included in the decision.