As wireless systems evolve toward 6G, AI becomes increasingly central to how networks operate, optimize themselves, and deliver advanced services. But with greater autonomy and complexity comes a critical requirement: the decisions made by AI models must be understandable, transparent, and trustworthy. This paper explores the technical foundations and emerging research challenges of Explainable Artificial Intelligence (XAI) in the context of 6G networks. In 6G systems, AI drives core functionalities—from resource allocation and network slicing to semantic communications and mission-critical applications. However, many high-performing AI models operate as “black boxes,” making it difficult for operators, users, and regulators to assess why a particular decision was made. XAI seeks to address this challenge by providing insights into model behavior, ensuring that network decisions remain interpretable even when they rely on complex learning pipelines.
The paper outlines several technical aspects of bringing XAI into 6G. These include designing interpretable models suitable for real-time operation, developing post-hoc explanation tools that shed light on opaque models, and ensuring that explanations remain meaningful across heterogeneous network components. The authors also emphasize the importance of evaluating explainability itself defining metrics, benchmarking methods, and studying how explanations influence trust and reliability in real deployments. A set of research challenges is also highlighted. Among them are the need to balance model performance with interpretability, adapt XAI methods to resource-constrained edge devices, integrate explainability into distributed and federated learning frameworks, and ensure that explanations remain robust against adversarial manipulation. These challenges illustrate that XAI is not simply an add-on to AI models but a fundamental design requirement for future 6G intelligence. Overall, the paper frames XAI as an essential pillar of 6G: a way to build transparent, reliable, and human-aligned network intelligence capable of supporting the next generation of wireless services.
Explainable_AI_for_6G_Use_Cases_Technical_Aspects_and_Research_Challenges