The lack of accountability in black-box Artificial

The gap

Evidence profile

Sourced from the inline gaps and future work and limitations and recommendations and stated research gap of the source papers, classified as general, spanning 4 journals.

Supporting evidence — 5 representative gaps

• Clustering and Classification Techniques in Machine Learning-based Intrusion Detection Systems: A Review (2026) · International Journal of Current Engineering and Technology · doi

  Future research should investigate explainable AI (XAI) techniques that can provide meaningful insights into the detection logic of ML- based IDS models. Future work should focus on the development of adaptive, lightweight, and hybrid IDS frameworks that can maintain high detection performance across diverse and evolving network environments while meeting the practical constraints of real-world deployment.

  generalinline gapsevidence 5/5
  Keywords: future detection investigate explainable techniques provide meaningful insights logic based models focus development adaptive lightweight
• AI-Enabled Intrusion Detection Framework for Secure Smart Network Environments (2026) · International Journal of Creative and Open Research in Engineering and Management · doi

  Future research can improve the proposed system by integrating deep learning techniques such as convolutional neural networks and recurrent neural networks. Additionally, implementing explainable AI models can improve transparency in intrusion detection decisions. Further studies can also focus on deploying the system in real-time smart network environments and improving detection performance for emerging cyber threats.

  generalfuture workevidence 5/5
  Keywords: improve system neural networks detection future proposed integrating deep learning techniques convolutional recurrent additionally implementing
• EXPLAINABLE AI-BASED INTRUSION DETECTION FRAMEWORK FOR CRITICAL INFRASTRUCTURE PROTECTION IN PAKISTAN'S DIGITAL ECOSYSTEM (2026) · Zenodo (CERN European Organization for Nuclear Research) · doi

  limited related cyber face to trust, related security to and tailored and systems ecosystem intelligent technologies. Existing cybersecurity systems within many organizations remain reactive rather than predictive and intelligent. Additionally, limited research has focused specifically on explainable AI-based intrusion detection to Pakistan’s digital critical infrastructure requirements. The reviewed literature indicates that although AI and ML-based frameworks have IDS significantly improved cybersecurity capabilities, transparency, challenges interpretability, false-positive reduction remain unresolved. Existing studies detection focused have performance while giving limited attention to explainability and contextual adaptation for developing as Pakistan.

  generallimitationsevidence 5/5
  Keywords: limited related systems intelligent existing cybersecurity remain focused explainable based detection pakistan cyber face trust
• EXPLAINABLE AI-BASED INTRUSION DETECTION FRAMEWORK FOR CRITICAL INFRASTRUCTURE PROTECTION IN PAKISTAN'S DIGITAL ECOSYSTEM (2026) · Zenodo (CERN European Organization for Nuclear Research) · doi

  It is recommended that organizations managing critical adopt infrastructure in Pakistan intrusion detection Explainable AI-based systems to improve cybersecurity resilience and operational transparency. Cybersecurity teams should be trained to interpret explainable AI outputs effectively to enhance incident response capabilities. Furthermore, policymakers should encourage into the national cybersecurity frameworks and promote research collaborations between academia, industry, institutions.

  generalrecommendationsevidence 5/5
  Keywords: cybersecurity explainable recommended organizations managing critical adopt infrastructure pakistan intrusion detection based systems improve resilience
• eXplainable Artificial Intelligence for Transparent Optimization of Deep Learning-Based Intrusion Detection Systems (2026) · Journal of Network and Systems Management · doi

  The lack of accountability in black-box Artificial Intelligence for IT Operations powered by Deep Learning. The need for a transparent conceptual model for the cyclical explanation and optimization of black-box Intrusion Detection Systems. The limitations of traditional rule-based methods of network monitoring and threat management.

  generalstated research gapevidence 5/5
  Keywords: lack accountability black-box artificial intelligence operations powered deep

Questions about this gap