The session delves into the imperative of AI integration in IMT-2030 (6G), emphasizing the need for AI-native design to enable new applications and services seamlessly. Unlike overlays or retrofitting, AI-native approaches ensure harmonious integration, where AI becomes intrinsic to user experiences. Discussions revolve around the transformation of networks into AI-native platforms, facilitating efficient data processing, AI algorithm execution, and application services. Key considerations include intelligent workload placement, energy-performance management, and privacy assurance, transitioning networks from AI-as-a-Service providers to integral AI-native platforms. Central to these discussions are the creation, configuration, and management of network functions autonomously, alongside enabling distributed edge nodes for inference. The session aims to identify challenges and opportunities in hosting AI as a native service within network architectures for IMT-2030 (6G).
