SolRugDetector: Investigating Rug Pulls on Solana

A research team from Sun Yat-sen University and other institutions has published a groundbreaking study on arXiv titled "SolRugDetector: Investigating Rug Pulls on Solana." The paper presents a comprehensive empirical analysis of fraudulent token schemes on the Solana blockchain, where the low barrier to token creation has led to widespread scams. Unlike Ethereum-based rug pulls that often involve malicious smart contracts, Solana's unified SPL Token program shifts fraudulent behavior to on-chain operations like market manipulation, creating new detection challenges that existing tools couldn't address.

The researchers first built a manually labeled dataset of 117 confirmed rug pull tokens based on 68 real-world incident reports. Using this data, they developed SolRugDetector, a detection system that identifies fraudulent tokens solely by analyzing on-chain transaction and state data. When tested, SolRugDetector outperformed existing detection tools on their labeled dataset. The team then conducted a large-scale measurement study of 100,063 tokens newly issued in the first half of 2025, identifying a staggering 76,469 (over 76%) as rug pulls.

Their analysis reveals that Solana rug pulls exhibit extremely short lifecycles, strong price-driven dynamics, severe economic losses, and highly organized group behaviors. The researchers have released both their detection system and the resulting dataset to support the development of more effective on-chain defense mechanisms. This work provides crucial insights into the scale and patterns of fraud on one of the fastest-growing blockchain platforms, where token creation costs are minimal but detection has been lacking.