This section summarizes selected research interests and ongoing methodological work across digital forensics, cybersecurity, AI robustness, OSINT, and external exposure assessment.
The main research activity concerns the operational robustness of AI-based tools in digital forensics. The thesis investigates how AI-assisted image classification behaves under clean conditions, adversarial perturbations, anti-forensic transformations, and out-of-distribution inputs.
Evaluation of AI-based systems in forensic-oriented workflows, with emphasis on reliability, traceability, explainability, and failure analysis.
Study of adversarial attacks and anti-forensic image transformations as stressors for AI-assisted digital evidence analysis.
Applied work on external signals, public-source intelligence, cyber-relevant monitoring, and situational awareness for investigative and security contexts.
Lightweight security assessment and vendor-risk screening concepts focused on domain-level signals, visible misconfigurations, and risk triage.