Explainable AI-based Intrusion Detection System for Industry 5.0: An Overview of the Literature, associated Challenges, the existing Solutions, and Potential Research Directions

Explainability is critical for cybersecurity trust: The paper notes that cybersecurity professionals have been "reluctant to accept black-box ML solutions," highlighting that in high-stakes security applications, understanding why an AI system flags a threat is essential for human decision-makers to trust and act on those alerts.

Industry 5.0's complexity creates new vulnerabilities: The increased interconnection of robots, IoT devices, AR/VR systems, and smart devices in manufacturing creates "several types of potential security flaws," meaning that as industries become more automated and AI-dependent, the attack surface for intrusions expands significantly.

Adversarial attacks threaten AI-based security systems: The paper examines "Adversarial XIDS approaches," indicating that malicious actors can specifically craft attacks to fool AI-based intrusion detection systems, meaning explainability becomes even more important for identifying when systems might be manipulated or compromised.

XAI bridges the gap between AI power and practical adoption: While AI has proven "very effective" at detecting intrusions, malware, and phishing, organizations won't deploy these systems at scale without understanding how they work—explainable AI addresses this adoption barrier in critical sectors like defense, healthcare, and finance.

Significant research gaps remain for Industry 5.0 security: The paper identifies "challenges and opportunities" requiring "future research," suggesting that current XAI-based intrusion detection solutions are not yet fully mature for the complex, interconnected environments of modern manufacturing.