This research introduces a novel framework leveraging large language models to detect illegal drug trafficking on online platforms despite highly imbalanced data.

• Addresses the critical challenge of class imbalance and limited labeled samples in real-world detection scenarios
• Employs graph prompt learning combined with LLMs to enhance detection accuracy
• Creates a more practical approach for identifying illegal drug activities on major online platforms
• Demonstrates how AI can support security and law enforcement in monitoring illicit online marketplaces

This innovation matters because it provides security professionals with more effective tools to identify and combat drug trafficking activities that increasingly operate through mainstream online channels.

LLM-Empowered Class Imbalanced Graph Prompt Learning for Online Drug Trafficking Detection