Published January 18, 2026 | Version v1.0
Preprint Open

Cerberus: Autonomous AI-Based Threat Detection and On-Chain Reputation Recording for U2U Networks

Authors/Creators

  • 1. ROR icon Gunadarma University

Description

Cerberus Watchdog is an autonomous threat detection framework designed to monitor the U2U
blockchain mempool in real time, identify potentially malicious transactions using artificial intelligence
techniques, and immutably record validated incidents on-chain through a dedicated smart contract. By
combining mempool-level surveillance, AI-assisted classification, and blockchain-based reputation
recording, Cerberus Watchdog provides early-warning capabilities, transparency, and accountability
within decentralized ecosystems. This paper presents the system architecture, design rationale,
implementation details, detection workflow, preliminary evaluation, security and ethical considerations,
limitations, and future research directions. The results indicate that AI-assisted mempool monitoring is
a feasible and promising approach for proactive blockchain security.

Files

Cerberus – Research Paper.pdf

Files (25.6 kB)

Name Size Download all
md5:0eeab3963c4bb653e78a9355ffe021cb
25.6 kB Preview Download

Additional details

Software

Repository URL
https://github.com/laudzakusuma/CERBERUS
Programming language
Python , JavaScript , Solidity
Development Status
Wip