AI-Enabled UAV Networks for 6G: A Paradigm Shift

Unmanned aerial vehicles (UAVs) are evolving from remotely controlled platforms to autonomous agents that extend the aerial layer of future 6G networks. This evolution is enabled by an integrated AI framework in which deep reinforcement learning (DRL) supports decision making, while graph neural networks (GNNs) capture dynamic network connectivity. Compared with conventional methods, AI-based approaches are more suitable for dynamic UAV environments with changing topology, interference, and security conditions. This article reviews recent progress in AI-enabled UAV networks, with an emphasis on the roles of DRL and GNNs in resource management and security. We also outline several open challenges, including multi-objective decision making, multi-UAV coordination, and hierarchical AI framework in space-air-ground integrated networks (SAGINs), and highlight future directions toward secure and intelligent UAV networks for 6G.