Intrusion Detection Systems (IDS) play a central role in security operations, yet analysts remain burdened by excessive false positives and opaque model behavior. Artificial Intelligence (AI) has the potential to improve detection rates, but many AI-driven IDS operate as black boxes that limit trust, usability, and accountability. This paper introduces the design and prototype of the Collaborative Intrusion Detection Interface and Interaction (CIDIX), a system that explicitly supports human–AI interaction in intrusion monitoring and response. CIDIX integrates automated detection, multi-level explainability, interactive visualization, and analyst-oriented recommendations to enhance transparency and shared decision-making. A proof-of-concept prototype demonstrates how analysts can interpret AI-generated alerts, explore attack graphs, and act on rules-based guidance. Although the current implementation is demonstrated with synthetic data, the case studies illustrate how explainable and visual interfaces help analysts build trust, reduce cognitive load, and strengthen collaboration with AI systems. This work advances IDS design beyond static alerts, positioning human–AI collaboration as a pathway toward shared intelligence in security operations.