Master's Thesis
A Human-in-the-Loop Security Control Validation Framework for Data Exfiltration Detection
Goal
Security teams struggle to know whether their detection controls actually work — and when they don't, whether it's a configuration problem, a gap in coverage, or a missed attack pattern. This thesis builds a framework that puts a human analyst in the loop: systematically stress-testing security controls against data exfiltration scenarios and surfacing where they need to be tuned.
Why it matters
Most detection validation happens reactively — after an incident. This framework flips that: it gives security teams a structured way to continuously probe their own defences, identify blind spots before attackers do, and build confidence that their controls hold under realistic conditions. The human-in-the-loop design means the results are actionable, not just another automated report.