Authors: Wenkai Li, Xiaoqi Li, Yuqing Zhang, Zongwei Li

Published on: May 17, 2024

Impact Score: 7.8

Arxiv code: Arxiv:2405.11035

Summary

• What is new: DeFiTail is the first to use deep learning to detect specific exploits in DeFi protocols.
• Why this is important: Attacks on DeFi protocols have resulted in significant financial losses, and current detection methods are inadequate.
• What the research proposes: The DeFiTail framework identifies access control and flash loan exploits using a deep learning model that processes unified transaction data paths.
• Results: DeFiTail shows high accuracy in detecting exploits with 98.39% for access control and 97.43% for flash loan, and 86.67% on identifying malicious contracts from CVE data.

Technical Details

Technological frameworks used: DeFiTail, leveraging deep learning

Models used: Deep learning models for detecting exploits

Data used: Control Flow Graph (CFG), symbolic execution stack for data validation

Potential Impact

Cryptocurrency and financial security sectors, potentially affecting companies involved in blockchain technology and crypto asset management.

Want to implement this idea in a business?

We have generated a startup concept here: DeFiShield.