The authors claim that a black-box attacker can identify sensitive traffic features from side-channel indicators and reduce an IDS's accuracy from 99% to 48% while staying invisible to an anomaly detector.
Generative adversarial attacks against intrusion detection systems using active learning,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CR 1years
2025 1verdicts
REJECT 1roles
background 1polarities
support 1representative citing papers
citing papers explorer
-
Vulnerability Disclosure through Adaptive Black-Box Adversarial Attacks on NIDS
The authors claim that a black-box attacker can identify sensitive traffic features from side-channel indicators and reduce an IDS's accuracy from 99% to 48% while staying invisible to an anomaly detector.