Smart contracts have been widely applied in various fields. Due to the immuta-bility of data on the blockchain, it is of great significance to conduct smart con-tract vulnerability detection before data is uploaded to the chain. To address the problems of low accuracy and single vulnerability type in traditional detection methods, a blockchain smart contract vulnerability detection method based on Graph Neural Network (GNN) is proposed. This method abstracts the functions and key code segments in smart contracts into nodes in a graph, and constructs edges by leveraging data and control dependencies during code execution, thereby accurately depicting the specific graph structures of reentrancy attacks and timestamp-dependent vulnerabilities. To further enhance the model’s sensi-tivity to key vulnerability patterns, the multi-head attention mechanism is in-novatively introduced, which can effectively screen out the nodes and edges that contribute the most to vulnerability detection, suppress irrelevant or noisy information, and significantly improve the accuracy and robustness of vulnera-bility detection. Experimental results show that the proposed method achieves an accuracy of 85.19% in reentrancy vulnerability detection and 82.37% in timestamp-dependent vulnerability detection, demonstrating excellent vulner-ability identification capability.
The rapid growth of blockchain technology and the increasing number of network nodes have heightened the risk of sophisticated attacks. Among these, Eclipse attacks present a serious threat to decentralized networks by exploiting their peer-to-peer structures. While previous research has explored artificial intelligence techniques to defend against Eclipse attacks, evolving attack patterns continue to challenge existing defenses. In this paper, we propose a novel defense framework that integrates a clustering approach based on self-attention encoders within a multi-kernel neural network clustering model. Our method utilizes parallel subnetworks to extract category-specific features from multiple perspectives, generating discriminative cluster centroids that are combined with raw transaction data to train a robust classifier for detecting Eclipse attacks in Ethereum networks. To evaluate our approach, we simulate Eclipse attacks on the Ethereum testnet and conduct extensive experiments. The results demonstrate that our method achieves a detection accuracy of 98.5% and improves classification performance by 5% compared to models trained without cluster-enhanced features, confirming the effectiveness of the proposed defense.
The pseudo-anonymity and rapidly expanding ecosystem of Decentralized Finance (DeFi) have brought about significant liquidity on EVM-compatible blockchains, making them lucrative targets for cybercriminals. In the modern financial landscape, the need for an automated, high-speed, and effective illicit money tracing system is more urgent than ever to support regulators, on-chain service providers and security practitioners in their efforts to combat the frequent and large-scale occurrences of cyber financial crimes. In this paper, we propose MFTracer, an automated system for tracing illicit money flows on EVM-compatible blockchains. Against the backdrop of a domain where tracing remains labor-intensive and expert-driven, MFTracer is developed in response to two pressing real-world demands: operational efficiency and forensic effectiveness. In response to the sophisticated fund transfer mechanisms enabled by the EVM environment, we introduce a novel fine-grained technique that enables protocol-agnostic transaction-level fund flow analysis. We further propose MFA, a lightweight and purpose-built graph abstraction with a tailored storage backend, to support efficient data retrieval. We also present a simulation algorithm for downstream illicit flow discovery. We implemented MFTracer. Its infrastructure for data retrieval achieves 3.7× to 9.4× higher storage efficiency while being 14.1× to 300× faster than the leading graph database systems. Furthermore, applied to real-world cybercrime incidents, MFTracer achieved 94.09% coverage of illicit money flows. It also newly reported 686 blockchain addresses and 4183 related transactions involved in money laundering that were previously undiscovered. MFTracer was able to reconstruct complete fund flow trajectories and provide strong evidence to investigators for 120.9 million in stolen assets.
Gukanraj S, Jeeva Rekha R, Dharshan M, Mohan Murthy M · 7 authors
Environmental pollution poses a significant threat to public health and ecosystems, demanding advanced methods for real-time monitoring and source identification. Traditional IoT monitoring systems often fail to capture complex spatiotemporal patterns and raise privacy concerns. This paper introduces a robust, privacy-preserving IoT-based environmental monitoring framework integrating Times Net for temporal feature extraction and Spatio-temporal Graph Neural Networks (STAGE) for spatial relationship modeling. The system incorporates Federated Learning with Differential Privacy, Zero-Knowledge Proofs (ZKP) for authentication, and Post-Quantum Cryptography (CRYSTALS-Cyber) for blockchain-secured model updates. Experimental evaluation using real-world IoT data demonstrates a 93.4% prediction accuracy, a 12% privacy gain, and a 35% reduction in communication cost compared to traditional methods. The architecture is scalable, modular, and designed to support real-time, privacy-sensitive environmental monitoring in smart city applications.
Graph-structured data has become central to modern analytics, enabling institutions to model relationships in domains such as healthcare, finance, cyber security, and education. However, privacy regulations and institutional policies restrict the sharing of sensitive nodes, edges, or interaction logs, preventing the discovery of global graph patterns. This paper introduces a novel framework for Federated Graph Pattern Mining Across Institutions (FGPM-AI), enabling multiple organizations to collaboratively extract global sub graphs, motifs, and temporal patterns without sharing raw graph data. The framework proposes six novel contributions: (1) Privacy-Preserving Pattern Signatures (PPPS) for anonymized sub graph encoding, (2) Federated Temporal Graph Pattern Mining (FT-GPM) to learn evolving patterns across distributed graphs, (3) Zero-Exchange Federated Sub graph Matching (ZE-FSM) using zero-knowledge proofs, (4) Heterogeneity-Aware Graph Pattern Consensus (HGPC) for semantic alignment between distinct graph schemas, (5) Communication-Adaptive Pattern Sharing (CA-FGM) for bandwidth-efficient collaboration, and (6) Multi-Party Graph Pattern Distillation (MGPD) for merging patterns into a unified knowledge model. Experimental design considerations demonstrate the feasibility and robustness of the framework. The results highlight FGPM-AI as a promising direction for secure, scalable, and intelligent cross-institution graph analytics.
Chunxia Liu, X T Zhu, Gaimei Gao, Junji Li · 5 authors
Security vulnerabilities in smart contracts pose serious threats to the trustworthiness and stability of blockchain systems. To overcome the limitations of existing detection methods in structural modeling and semantic feature representation, this paper proposes a smart contract vulnerability detection method (HEVD) that integrates heterogeneous graph attention networks with expert knowledge. HEVD achieves dynamic synergy between structural features and semantic priors through a multi-head self-attention mechanism. The heterogeneous graph attention network captures deep structural dependencies in smart contract code, while a hierarchical expert knowledge system distills high-confidence vulnerability patterns to provide explicit semantic guidance. This design effectively addresses the semantic gap of models that rely solely on data-driven training and compensates for the lack of global structural awareness in traditional expert-based approaches. Experimental results show that HEVD attains F1scores of 91.03%, 90.48%, and 83.46% for reentrancy, timestamp dependency, and infinite loop vulnerabilities, respectively, significantly outperforming multiple existing methods. Ablation studies further confirm the effectiveness of the proposed framework, demonstrating that the deep fusion of structural representation and semantic priors is key to enhancing smart contract vulnerability detection.
Large Language Models (LLMs) are increasingly deployed in high-stakes financial domains, yet they suffer from specific, reproducible hallucinations when performing arithmetic operations. Current mitigation strategies often treat the model as a black box. In this work, we propose a mechanistic approach to intrinsic hallucination detection. By applying Causal Tracing to the GPT-2 XL architecture on the ConvFinQA benchmark, we identify a dual-stage mechanism for arithmetic reasoning: a distributed computational scratchpad in middle layers (L12-L30) and a decisive aggregation circuit in late layers (specifically Layer 46). We verify this mechanism via an ablation study, demonstrating that suppressing Layer 46 reduces the model's confidence in hallucinatory outputs by 81.8%. Furthermore, we demonstrate that a linear probe trained on this layer generalizes to unseen financial topics with 98% accuracy, suggesting a universal geometry of arithmetic deception.
Blockchain technology has emerged as a transformative paradigm for decentralized systems, enabling secure, transparent, and tamper-resistant data management through distributed consensus mechanisms that eliminate the need for centralized control. At the core of these systems, consensus protocols ensure agreement among network participants; however, traditional approaches such as Proof of Work (PoW), Proof of Stake (PoS), and Byzantine Fault Tolerance (BFT) face persistent challenges related to scalability, energy consumption, and latency. In response, graph-theoretic approaches have gained prominence as an effective framework for modeling and optimizing blockchain consensus by representing nodes as vertices and communication links as edges, thereby capturing complex network relationships, trust structures, and interaction patterns. This paper systematically reviews graph-based methods applied to blockchain consensus, highlighting their role in improving efficiency, enhancing security against attacks such as Sybil and double-spending, and optimizing node selection. Advanced techniques including graph partitioning, spectral clustering, and network flow optimization further contribute to improved scalability and throughput. The study identifies a clear transition toward intelligent, hybrid consensus mechanisms integrating graph theory, machine learning, and distributed computing, while also addressing ongoing challenges such as computational complexity and dynamic adaptability, and outlining future directions for AI-driven, scalable, and secure consensus models.
Zero-knowledge proofs of training (zkPoT) allow a party to prove that a model is trained correctly on a committed dataset without revealing any additional information about the model or the dataset. Existing zkPoT protocols prove the entire training process in zero knowledge; i.e., they prove that the final model was obtained in an iterative fashion starting from the training data and a random seed (and potentially other parameters) and applying the correct algorithm at each iteration. This approach inherently requires the prover to perform work linear to the number of iterations.
Smart contracts are self-executing programs on blockchains, critical for enabling efficient, secure, and reliable data exchange and value transfer. However, as their application scenarios expand, reliability issues have become a major bottleneck for blockchain development. Existing vulnerability detection methods often model smart contract source code as graph structures and use Graph Neural Networks (GNNs) for feature learning. Yet these methods over-rely on static execution flow features and ignore dynamic behavioral information of contract accounts in real runtime environments, limiting their ability to capture dynamic patterns and semantic details of contracts. To address these challenges, this paper proposes a graph attention network vulnerability detection model with global feature augmentation for smart contracts (GaGAT). Specifically, we first model key functions and variables in the contract source code as nodes, and execution flows as edges to construct a base contract graph. Then, we innovatively introduce global virtual nodes that integrate two types of information: contract categories and contract account behavioral features, including balance changes, Ether inflow/outflow, daily transaction frequency and single transaction duration. After contract graph contraction and feature transformation, we generate a feature matrix as input to the GaGAT model. Subsequently, we conduct vulnerability detection. Through a series of experiments, we provide empirical evidence of the superior performance of our proposed method compared to existing approaches in detecting six different categories of vulnerabilities.This study provides a new paradigm for cross-modal feature fusion for smart contract security analysis.
Abstract Local Energy Communities (LECs) are gaining prominence as key actors in the transition toward sustainable and decentralized energy systems. A critical challenge for these communities lies in achieving energy self-sufficiency through effective forecasting of energy production and consumption. Accurate forecasting models are essential to support optimization and planning strategies. However, privacy concerns and regulatory constraints often limit the feasibility of centralized data-driven approaches, as users are understandably reluctant to share their consumption data. To address this issue, we propose a privacy-preserving forecasting framework based on Federated Learning (FL) and Long Short-Term Memory (LSTM) networks, which enables collaborative model training without disclosing raw user data. Building upon this core architecture, we further enhance transparency and user engagement by introducing Zero-Knowledge Proofs (ZKPs) for secure inference verification, and a novel incentive layer based on dynamic Non-Fungible Tokens (dNFTs) and fungibile tokens. Our approach ensures model integrity, protects user data, and fosters sustainable behavior through verifiable, trustless reward mechanisms. Experimental results demonstrate the feasibility and potential of this architecture in supporting privacy-aware, decentralized energy forecasting within LECs.
Graph learning has garnered increasing attention in recent years, which aims to train machine learning models over graph data to support various graph analytic tasks. Coming with the popularity of graph learning are critical privacy concerns regarding the information-rich graphs in many application domains (e.g., finance, social networks, and healthcare). There is thus an urgent call for privacy-preserving graph learning. In this paper, we target an emerging decentralized graph scenario, where a graph is fully decentralized among a set of nodes in such a way that each node only has a limited local view about the global graph. We propose PDGL, a new system framework that can effectively support privacy-assured model training over a decentralized graph, with privacy protection for the links among the nodes as well as the nodes’ private feature data and labels. In contrast to PDGL, prior work does not provide protection for the nodes’ links, feature data, and labels simultaneously. Extensive experiments demonstrate that while providing strong privacy protection for decentralized graph data, PDGL can achieve model utility comparable to the baseline setting of centralized graph learning.
Fraudulent activity on blockchain networks poses significant risks to the integrity and trust of decentralized finance ecosystems. The timely and accurate detection of fraud nodes such as phishing addresses within large-scale Ethereum transaction networks remains a major challenge due to their dynamic, sparse, and evolving structures. While methods like graph deep learning (e.g., graph neural networks) have been extensively explored, they are not inherently designed to capture higherorder interactions and textual information embedded within graph data. Motivated by the urgent need for advanced and robust fraud detection techniques, we introduce a novel graph prompting method named Large Language Model-Simplicial Complex (LLM-SC) based graph prompting framework that leverages LLM-based multi-agent collaboration system, LLMbased financial news prompt function, and simplicial neural networks to capture both the structural and contextual dimensions of blockchain activity. The empirical studies demonstrate the effectiveness of our approach, and these results provide a new tool for blockchain analytics platforms and regulatory authorities, enabling earlier and more accurate identification of fraudulent behavior and ultimately supporting safer and more resilient digital financial systems. The code is available at https://github.com/y13564/LLM-SC.
Because of the rapid acceleration of cloud computing, data transfer security and intrusion detection in cloud networks have become emerging areas of concern. All traditional security mechanisms have central vulnerabilities, cannot detect real-time threats, and are ineffective against zero-day attacks. Signature-based approaches of existing intrusion detection systems (IDS) do not cover the dynamically changing nature of cyber threats. Conventional blockchain security methods suffer from poor scalability and dynamic threat analysis. Therefore, this research proposes integrating Ethereum Blockchain and Deep Learning to construct a well-founded security framework for cloud networks with data migration security and real-time intrusion detection. The architecture has five distinct methods, each of which deals with particular security issues. Blockchain-Aware Federated Learning for Secure Model Training (BAFL SMT) guarantees tamper-proof and decentralized deep learning model training, which reduces model poisoning attacks by 98.4%. Graph Neural Networks for Adaptive Intrusion Detection (GNN-AID) captures graph structures for real-time anomaly detection in networks while reducing false positives to 1.2%. Quantum-inspired Variational Autoencoders (QI VAE ZDAD) provide enhanced zero-day attack detection, with an improved detection rate of 92%. Self-Supervised Contrastive Learning for Blockchain Security Auditing (SSCL-BSA) detects smart contract vulnerabilities automatically, resulting in an 87% reduction in fraud risk. Finally, Hierarchical Transformers for Secure Data Migration (HT SDM) enhance the transfer security of large-scale cloud data, achieving an attack classification accuracy of 99.1%. Overall, this multi-layer security framework will greatly enhance cloud security by preserving data integrity, cutting down the intrusion detection time by up to 65%, and enhancing response mechanisms. By marrying the immutable transparency of blockchain with superior anomaly detection at deep learning, this research provides a scalable, real-time, and intelligent approach to strengthening security against the backed-up transfer of data within cloud networks.
Ghislain Nkamdjin Njike, Anh-Tu Hoang, Stefan Schulte
Distributed learning (DL) is gaining popularity as it enables clients (e.g., AI Agents) to enhance their machine learning (ML) models’ performance by exchanging knowledge without revealing private datasets. State-of-the-art DL approaches primarily focus on transferring knowledge between heterogeneous clients with diverse model architectures, connecting clients with those that can improve their models, and protecting data privacy. However, they overlook the threat of malicious clients that potentially downgrade the models’ performance by sharing inaccurate knowledge or excluding high-performing clients from the training procedure.Therefore, we introduce the Zero-Knowledge Blockchain-Based Knowledge Distillation Learning Framework (zkBKD). In zkBKD, heterogeneous clients communicate with a blockchain network to discover high-performing clients, verify zero-knowledge proofs (ZKPs) to ensure the correctness of the knowledge shared from other clients, and vote to eliminate malicious clients. We analyze security and privacy risks and show that zkBKD prevents membership, poisoning, and collusion attacks. We conduct extensive experiments on two standard datasets across heterogeneous clients with four model architectures. The experimental results demonstrate that zkBKD relatively improves the average model accuracy of all clients by 25.71%. Even lightweight models such as ResNet-2 achieve up to a 103.35% accuracy gain compared to independent training.
Tuan-Dung Tran, Dinh Khang Nguyen, Quang Trung Do, Van-Hau Pham
Cross-chain bridges, while critical for interoperability in the Web3 ecosystem, have become a primary target for exploits, accounting for over $4.3 billion in losses—nearly 40% of all value stolen in recent years. Existing security paradigms fail to address this threat adequately due to a fundamental trade-off between pre-deployment static analysis and real-time dynamic monitoring. Current security paradigms are trapped in a critical trade-off: pre-deployment static analysis lacks runtime context and suffers from high false-positives, while real-time dynamic monitoring is blind to the underlying source-code vulnerabilities that enable sophisticated attacks. This paper introduces VeriBridge, a novel framework that breaks this impasse. VeriBridge pioneers a synergistic fusion of static intelligence and dynamic graph learning. It enriches real-time transaction graphs with fine-grained vulnerability data extracted from static analysis, providing crucial security context to an unsupervised Graph Autoencoder. By learning a high-fidelity model of normal behavior, VeriBridge detects malicious transactions, including zero-day exploits, as significant deviations from this learned norm, identified by high reconstruction error. Evaluated on a comprehensive dataset of real-world attacks, including the Poly Network and THORchain exploits, VeriBridge achieves a 96% F1-score, demonstrating a new frontier in robust, real-time security for critical blockchain infrastructure.
Aiming at the problem that smart contract security vulnerability detection faces high false positives in static analysis and low efficiency in dynamic analysis, which leads to low accuracy of vulnerability detection, a smart contract vulnerability detection method based on improved graph neural network (EGN, Event-Enhanced GNN) was proposed. Firstly, security mode features and graph features were extracted. Security mode features reduced false positives caused by blind detection, and graph features avoided the performance bottleneck of full graph traversal. Secondly, the high-risk functions were screened based on the risk probability threshold to improve the overall analysis efficiency. Thirdly, the temporal graph neural network was deployed for high-risk functions, the event temporal graph was dynamically tracked and the self-attention mechanism was used to capture vulnerabilities, so as to enhance the detection ability of complex vulnerabilities. Finally, we focus on reentrant vulnerability and timestamp dependency vulnerability detection. Through the evaluation experiments on the real contract datasets of two platforms of Ethereum and VNT chain, the experimental results show that the accuracy and F1 value of the proposed model for detecting reentries vulnerability reach 93.12% and 94.29% respectively, and the accuracy and F1 value of timestamp dependency vulnerability reach 91.71% and 91.42% respectively, which are better than the existing methods.
Smart contract vulnerability detection is an important task in securing the blockchain. However, existing detection methods primarily extract single view features, such as semantic or structural features, which ignores the synergistic supplementation of them to smart contract, remaining room for improvement in feature representation. To this end, this paper proposes the LLM-assisted dual-view awareness framework for smart contract vulnerability detection, which incorporates significantly different semantic features and structural features. To address the limitation of large language model (LLM) in domain-specific expertise, we design semantic awareness module based on Retrieval-Augmented Generation (RAG), construct vulnerability knowledge base, and perform semantic reasoning on smart contracts. To capture crucial structural information, we propose structural awareness module based on Graph Neural Network (GNN), construct contract graphs, and perform structural analysis on smart contracts. We evaluated four types of vulnerabilities, and the experimental results show that our approach significantly outperforms state-of-the-art approaches, achieving 4.80% improvement in accuracy for timestamp dependence detection.
Smart contracts control critical financial assets on blockchains, with potential weaknesses risking substantial losses. Thus, smart contract vulnerability detection is essential for maintaining blockchain ecosystem stability. Traditional methods depend extensively on expert-driven patterns, resulting in poor scalability. Although deep learning-based approaches have made significant progress, they still suffer from issues such as inflexible representations, insufficient feature modalities, and limited model capabilities. In this paper, we propose FSGDec, a novel smart contract vulnerability detection framework that fuses sequential information and structural features at the bytecode level. Firstly, an efficient node embedding method is developed for contract control flow graphs, flexibly processing node sequences and incorporating node-specific semantic information associated with weaknesses. Then, by modeling node features as time series signals, an adaptive graph wave network is introduced to automatically capture vulnerability-related structural features. Finally, a classifier is deployed to perform bug detection utilizing the extracted graph-level features that integrate semantic information. Evaluated on two real-world smart contract datasets, the experimental results demonstrate that FSGDec achieves superior performance compared to state-of-the-art baselines.
Love Allen Chijioke Ahakonye, Cosmas Ifeanyi Nwakanma, Jae Min Lee, Dong‐Seong Kim
The growing complexity of distributed industrial IoT systems heightens cybersecurity risks, exposing the limitations of centralized ML-based intrusion detection. Federated Learning (FL) enables decentralized, privacy-preserving model training but remains susceptible to adversarial threats and system-level failures. This study introduces PureChain, a decentralized ledger using a proof-of-authority and association (PoA 2 ) consensus mechanism to enhance FL-based IDS security. The study offers insight into the mathematical model of the PureChain-enhanced FL, which integrates blockchain-inspired consensus protocols for collaborative intrusion detection across organizations, ensuring data privacy while providing tamper-proof logs and automated responses through smart contracts. It incorporates dynamic fault tolerance, poisoning resistance, and privacy preservation with FL, enhancing security and performance in decentralized systems. Experimentation with varying client subsets demonstrates its adaptability with a TPS range of 312 . 5 − 1178 . 3 and a low latency range of 0 . 0008484 − 0 . 0032 . The framework ensures comprehensive security, reliability, and privacy, providing a scalable solution for decentralized, secure systems.
With the rise of cryptocurrencies, illicit activities such as money laundering, fraud, and Ponzi schemes have gained attention. Traditional methods using graph neural networks (GNNs) to detect illicit transactions treat the entire transaction network as input, which works well on small networks but struggles with large-scale blockchain data. To address this limitation, the authors propose a neighborhood subgraph-based method that combines GCN and LSTM. The GCN captures information from neighboring nodes for each transaction, enhancing the understanding of the network structure, while the LSTM tracks the sequence and variations of fund flows. Experimental results show that by using 3-hop neighborhood subgraphs, the method outperforms other baseline models while requiring data from only an average of 80 nodes, thereby significantly improving efficiency compared to methods that process the entire transaction network.