AI IN OFFENSIVE CYBERSECURITY: AUTONOMOUS TTP EMULATION AND SCALABLE CONTINUOUS TESTING

Sali Saliev
Additional contact information
Sali Saliev: University of National and World Economy, Sofia, Bulgaria

Innovative Information Technologies for Economy Digitalization (IITED), 2025, issue 1, 42-53

Abstract: Offensive testing is shifting from ad-hoc, manual exercises to routine, automated checks. We present a system that uses large language models for planning and tool-driven agents for controlled TTP execution against corporate-like lab environments. The approach spans reconnaissance, initial access, lateral movement, and impact while operating under explicit risk policies, scope limits, and an emergency stop. A built-in telemetry pipeline turns technical findings into management-ready exposure metrics (asset criticality, blast radius, attack-path length). In lab scenarios, the system delivers results comparable to human red teams and reduces time-to-finding by 35–60% depending on the case, while enabling nightly "attack rehearsals" to track security drift. We detail prompt-engineering tactics, synthetic-data hardening, and countermeasures against hallucinations, scope creep, and OPSEC leaks. The outcome is a pragmatic “human intent + machine scale†model for continuous control validation and decision-ready analytics.

Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
https://www.unwe.bg/doi/iited/2025/IITED.2025.05.pdf (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:nwe:iitfed:y:2024:i:1:p:42-53

Access Statistics for this article

More articles in Innovative Information Technologies for Economy Digitalization (IITED) from University of National and World Economy, Sofia, Bulgaria Contact information at EDIRC.
Bibliographic data for series maintained by Vanya Lazarova ().