Recent Advances in Generative Artificial Intelligence for Communication Networks

Generative Artificial Intelligence (GenAI) is set to transform networking by enabling new approaches to management, security and privacy. By learning patterns from existing data and generating configurations, traces and policies, GenAI enhances automation, optimisation, troubleshooting and fault resilience. Applications range from predictive modelling and natural language-based orchestration to anomaly detection, configuration optimisation and intelligent script generation. At the same time, challenges remain in data quality, explainability, robustness and compliance with privacy regulations. This paper surveys recent advances in applying GenAI to networking, structured around unifying operational patterns (Intent?Plan?Act, Predict?Allocate?Adapt, Detect?Explain?Recover). We analyse state-of-the-art techniques in network management and security, highlight their limitations and outline future research directions toward telecom-grade, resilient and trustworthy GenAI-enabled infrastructures. © 2020 IEEE.