Frontiers of Generative AI for Network Optimization: Theories, Limits, and Visions

While interest in the application of generative AI (GenAI) in network optimization has surged in recent years, its rapid progress has often overshadowed critical limitations intrinsic to generative models that remain insufficiently examined in existing literature. This survey provides a comprehensive review and critical analysis of GenAI’s role in network optimization. We focus on the two dominant paradigms of generative models, including generative diffusion models (GDMs) and large pre-trained models (LPTMs), and organize our discussion around a categorization we introduce, dividing network optimization problems into two primary formulations: one-shot optimization and Markov decision process (MDP). We first trace key developments, including foundational contributions from the broader AI community, and systematically categorize current efforts in networking. We also briefly review frontier applications of GDMs and LPTMs in other related networking tasks, providing additional context for this survey. Building on this, we present theoretical generalization bounds for GDMs in both one-shot and MDP settings. The established bounds also enable an equivalent analysis of LPTMs. Most importantly, we reflect on the overestimated perception of GenAI’s general capabilities and caution against the all-in-one illusion it may convey. We highlight critical limitations, including difficulties in satisfying hard constraints, limited concept understanding, and the inherently probabilistic nature of outputs. We also propose key directions for future research, such as bridging the gap between generation and optimization. Although these two tasks are increasingly integrated in practical applications, they differ fundamentally in both objectives and underlying mechanisms, necessitating a deeper understanding of their theoretical connections. Ultimately, this survey aims to provide researchers and practitioners with a structured overview and a deeper insight into the strengths, limitations, and potential of GenAI in network optimization.