Agents Should Replace Narrow Predictive AI as the Orchestrator in 6G AI-RAN

This aggregated arXiv feed contains multiple network- and ML-focused papers with notable security and privacy implications. Key points: (1) “Agents Should Replace Narrow Predictive AI as the Orchestrator in 6G AI‑RAN” argues for LLMs/LTMs as RAN orchestrators in the RIC and explicitly calls out the need for continuous alignment (RLNF), neuro-symbolic verification to reduce hallucinations, extreme quantization for edge deployment, and hardening orchestration frameworks against adversarial/prompt‑injection attacks — introducing new attack surface and supply-chain/control risks for mobile-network