ExplainabilityTransparencyTrustworthinesscs.CR
Explainable Artificial Intelligence and Cybersecurity: A Systematic Literature Review
Carlos Mendes, Tatiane Nogueira Rios
February 27, 2023
arXiv.org
24 citations
1 influential
Key Takeaways
-
Black-box AI creates trust barriers in cybersecurity: The paper identifies that traditional AI techniques like deep neural networks make decisions that cybersecurity operators cannot understand or verify, creating a significant adoption challenge in security-critical environments where human oversight is essential.
-
Explainable AI (XAI) bridges the comprehension gap: XAI techniques are specifically designed to make AI decision-making transparent and interpretable, addressing the core problem that prevents cybersecurity professionals from confidently deploying and monitoring AI-powered security solutions.
-
XAI adoption remains inconsistent across cybersecurity domains: The systematic review reveals that while AI is widely applied across cybersecurity, the integration of explainability techniques is uneven—suggesting some security areas have benefited from XAI while others continue relying on opaque AI systems.
-
Human understanding is critical for security operations: The paper emphasizes that cybersecurity operators need to comprehend why an AI system flagged a threat or made a security decision, highlighting that responsible AI deployment in security requires more than just accuracy—it requires interpretability that enables human judgment and accountability.
Abstract
Cybersecurity vendors consistently apply AI (Artificial Intelligence) to their solutions and many cybersecurity domains can benefit from AI technology. However, black-box AI techniques present some difficulties in comprehension and adoption by its operators, given that their decisions are not always humanly understandable (as is usually the case with deep neural networks, for example). Since it aims to make the operation of AI algorithms more interpretable for its users and developers, XAI (eXplainable Artificial Intelligence) can be used to address this issue. Through a systematic literature review, this work seeks to investigate the current research scenario on XAI applied to cybersecurity, aiming to discover which XAI techniques have been applied in cybersecurity, and which areas of cybersecurity have already benefited from this technology.