Cyber risk quantification for adversarial machine learning attacks

doi:10.1016/j.compeleceng.2026.110964

Cyber risk quantification for adversarial machine learning attacks

2026-01-09

0

PRE

AI

DOI：10.1016/j.compeleceng.2026.110964

Original

Original request for help

Save

Abstract

En 中文

• Unified risk framework integrating Bow-Tie, MITRE ATT&CK, FAIR and Monte Carlo. • The approach maps attacker paths and defense controls to clarify risks for technical stakeholders. • Probabilistic risk modeling estimates annualized loss to support data-driven security decisions. • Case study on a hypothetical ML ransomware attack quantifies organizational impact. • The framework extends its scope to evasion, poisoning, and privacy attacks.

Cyber risk quantification for adversarial machine learning attacks

Abstract

Journal

Organization