Artificial Intelligence as a Service (AIaaS) enables users to query a model hosted by a service provider and receive inference results from a pre-trained model. Although AIaaS makes artificial intelligence more accessible, particularly for resource-limited users, it also raises verifiability and privacy concerns for the client and server, respectively. While zero-knowledge proof techniques can address these concerns simultaneously, they incur high proving costs due to the non-linear operations involved in AI inference and suffer from precision loss because they rely on fixed-point representations to model real numbers.
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.
Smart contracts are commonly audited through static analysis to explore vulnerabilities. However, static approaches typically produce heterogeneous findings rather than reproducible, executable proof-of-concept (PoC) test cases, leading to costly and ad hoc manual validation. Large language models (LLMs) offer a promising way to translate audit reports into PoC test cases, but face three major challenges: noisy inputs, lack of execution grounding, and missing runtime oracles. We present SmartPoC, an end-to-end approach for validating reported vulnerabilities in audit reports by generating and executing PoC test cases with automated exploitability verification. SmartPoC first extracts a focused function-level slice from each report to reduce noise, centering on the key functions referenced in a finding and augmenting them with execution-relevant neighbors. To improve executability, we wrap LLM-based PoC synthesis in a generate-repair-execute loop, combining deterministic pre-execution sanitization with feedback-driven post-execution debugging. We further use differential verification as an oracle to confirm the exploitability of generated test cases. On the SmartBugs-Vul and FORGE-Vul benchmarks, SmartPoC achieves confirmation precision of 98.32% and 98.65%, with recall of 84.17% and 85.28%, respectively. On a recent Etherscan verified-source corpus, SmartPoC confirms 64 bugs from 545 audit findings at an average cost of $0.03.
Smart contracts are a core component of blockchain ecosystems, but their transparency and immutability make them vulnerable to attacks, leading to significant financial losses. Thus, repairing vulnerabilities in smart contracts is crucial for establishing a trustworthy blockchain environment. Existing smart contract vulnerability repair methods suffer from a critical "one-for-all" design limitation, where a single model is tasked with fixing diverse vulnerability types, leading to suboptimal performance due to insufficient specialization. To address this, we propose MoEFix, a novel framework leveraging a Mixture-of-Experts (MoE) architecture tailored for smart contract characteristics. MoEFix partitions vulnerabilities into subspaces, trains specialized experts for each type (e.g., reentrancy, integer overflow), and employs a vulnerability-aware router to dynamically allocate repairs. We further redesign the repair workflow to align with large language models, enabling end-to-end secure contract generation instead of partial patches, and to achieve this, we curated a dataset of 1,391 contracts covering five critical vulnerability types.To validate our approach, we extend the benchmark PVD test suite. Experiments demonstrate that MoEFix outperforms state-of-the-art methods by 21.64% in overall accuracy, achieving improvements of 26.19% (reentrancy) and 23.08% (delegatecall) for specific vulnerabilities.
Vulnerability attacks targeting smart contracts have caused significant losses of digital assets. Many approaches based on static analysis, fuzzing, and deep learning have been proposed for detecting contract vulnerabilities. However, most existing methods only support vulnerability detection within individual contracts. When contracts interact with each other through external calls, these methods fail to perform effective cross-contract security analysis, leading to false negatives and false positives. To address these limitations, we propose CrossMiner, a deep learning-based approach for vulnerability detection in contract interaction scenarios. CrossMiner enables comprehensive risk assessment for cross-contract security through trace analysis of function call chains. Specifically, CrossMiner first constructs a cross-contract dependency graph based on function call chains to effectively model inter-contract dependencies and network dynamics, and collect semantic information about contract interactions. Then, it employs a heterogeneous graph neural network with a two-level attention mechanism to finely extract and integrate complex features from the dependency graph, ultimately achieving precise risk assessment and vulnerability detection. We evaluate the effectiveness of CrossMiner on three types of smart contract vulnerabilities: reentrancy, timestamp dependency, and transaction state dependency. Experimental results demonstrate that CrossMiner achieves the best performance among all baseline methods, improving detection accuracy by 5.52%, 4.94%, and 5.60% for these vulnerabilities, and the F1 scores are improved by 5.44%, 5.02%, and 5.40%, respectively.
The rise of collaborative AI, particularly in distributed Mixture-of-Experts (MoE) systems, has created a critical challenge: how to ensure trust and transparency when aggregating proprietary models from different providers. To address this, we introduce a novel cryptographic protocol ZQ-WMA that enables verifiable and privacy-preserving online learning. Our method integrates zero-knowledge proofs with a quantized version of the Weighted Majority Algorithm, allowing a central aggregator to publicly prove it is honestly combining expert advice and updating weights according to the agreed-upon rules, all without revealing any confidential model parameters.This approach ensures that expert contributions are evaluated fairly and protects valuable intellectual property. Our analysis reveals that the quantization necessary for the zero-knowledge proofs can counter-intuitively enhance prediction accuracy, a phenomenon we attribute to the maximal entropy random walks. Furthermore, our benchmarks demonstrate the efficiency of this method, showing proof generation complexity less than 10% of a standard SHA256 hash function, with O(1) proof size and verification time. This work provides a practical and scalable framework for building trustworthy collaborative AI systems.
Multi-agent systems (MAS) have emerged as a critical paradigm for distributed problem-solving in complex environments. However, their deployment in mission-critical applications faces significant challenges regarding trust, security, and adversarial robustness. This paper presents TrustOrch, a novel dynamic trust-aware orchestration framework designed to enhance the resilience of multi-agent collaboration against adversarial attacks. TrustOrch introduces five key innovations: (1) a dynamic trust assessment mechanism that evaluates agent reliability in real-time using multi-dimensional metrics, (2) an adversary-aware orchestration strategy combining reinforcement learning and game theory to detect and mitigate prompt injection attacks, (3) an adaptive collaboration topology that dynamically adjusts agent communication structures based on task complexity and trust levels, (4) explainable decision tracing for complete audit chains, and (5) a layered security architecture leveraging blockchain technology for decentralized trust verification. Our experimental evaluation demonstrates that TrustOrch reduces collision rates by 62%, achieves 91.7% robustness under adversarial attacks, and reduces communication overhead by 39.8% compared to baseline approaches. The framework achieves robust performance under various adversarial scenarios while maintaining transparency and regulatory compliance, making it particularly suitable for deployment in high-risk domains such as finance, healthcare, and autonomous systems.
J. Wenzel, Alam, Syeda Umaima, Andreas Schmidt, Hanwei Zhang · 5 authors
An ever increasing number of high-stake decisions are made or assisted by automated systems employing brittle artificial intelligence technology. There is a substantial risk that some of these decision induce harm to people, by infringing their well-being or their fundamental human rights. The state-of-the-art in AI systems makes little effort with respect to appropriate documentation of the decision process. This obstructs the ability to trace what went into a decision, which in turn is a prerequisite to any attempt of reconstructing a responsibility chain. Specifically, such traceability is linked to a documentation that will stand up in court when determining the cause of some AI-based decision that inadvertently or intentionally violates the law. This paper takes a radical, yet practical, approach to this problem, by enforcing the documentation of each and every component that goes into the training or inference of an automated decision. As such, it presents the first running workflow supporting the generation of tamper-proof, verifiable and exhaustive traces of AI decisions. In doing so, we expand the DBOM concept into an effective running workflow leveraging confidential computing technology. We demonstrate the inner workings of the workflow in the development of an app to tell poisonous and edible mushrooms apart, meant as a playful example of high-stake decision support.
Zero-knowledge proofs (ZKPs) have been used to protect the integrity of machine learning (ML) models. However, existing ZKPs for ML are still inefficient, mainly due to the computational cost of evaluating non-linear functions. In this paper, we propose a ZKP framework for typical non-linear functions in ML, including Sigmoid, Softmax, etc. Compared to the state-of-the-art Hao et al. (USENIX Security ’24), our protocols obtain 115.6-2384.4× and 296.8-4104.7× runtime improvements for prover and verifier, respectively, with a 37.91269.5× reduction in proof size.
With the deep integration of 6G, the Internet of Things, and artificial intelligence, this paper proposes an intrusion detection and defense framework that combines robust AI kernel reconstruction, a cross-layer collaborative perception architecture, and a dynamic defense closed-loop mechanism to address advanced persistent threats and dynamically evolving attacks targeting next-generation consumer services. First, a lightweight detection model ATF-KDBC is designed based on adversarial training and online knowledge distillation. Gradient masking and noise injection are employed to enhance robustness against adversarial samples, while a drift-aware module enables adaptive optimization under concept drift scenarios. The model achieves accuracies of 99.25% and 99.84% on the NSL-KDD and IoT-23 hybrid datasets, respectively, and compresses the model size to 1.08 MB, representing a 97.6% reduction compared with the BERT teacher model. Second, a multidimensional attack chain analysis model is developed based on a STHGN. By integrating semantic, structural, and temporal features with a multi-head self-attention mechanism, the model enables cross-layer threat tracing and millisecond-level response, achieving an F1-score exceeding 97.0% on the DARPA dataset. Furthermore, this study explores the construction of a distributed CTIS network by integrating federated learning and blockchain technology. Zero-knowledge proofs are employed to ensure privacy preservation, while a Quality of Data and Quality of Model scoring mechanism enables efficient and precise deployment of defense strategies. Experimental results demonstrate that the proposed framework significantly outperforms traditional methods in terms of robustness, environmental adaptability, and computational efficiency, thereby providing both theoretical support and a technical pathway for enhancing the resilience and security of next-generation consumer services.
Ebuka Chinaechetam Nkoro, Love Allen Chijioke Ahakonye, Dong‐Seong Kim
Smart Contracts (SCs), which are the backbone of automated transactions and digital assets within the Metaverse, ironically suffer from their own share of security vulnerabilities. While detecting these SC vulnerabilities using Artificial Intelligence (AI) and Deep Neural Networks (DNNs) has demonstrated remarkable performance and gained wide adoption, a critical limitation remains: the lack of explainability in these black box models. To facilitate meaningful progress in this field, our study addresses this gap by introducing a model-agnostic explanation framework that is both visual and quantitative, with human stakeholders actively involved to govern, verify, and interpret SC model predictions. The explainable SC outputs can be utilized for reward issuance and digital assets governance in the Metaverse. The effectiveness of our proposed Explainable AI (XAI) approach is validated using benchmark datasets, BCCC SCsVul 2024 and BCCC SCsVul 2023, comprising Ethereum SC entropy source codes, where it achieves an optimal detection accuracy of 97.13% alongside comprehensive explainability. To the best of our knowledge, this represents the first attempt at making Ethereum SC vulnerability detection within the Metaverse explainable, offering a valuable foundation for blockchain researchers, Metaverse security experts, and practitioners seeking verifiable, trustworthy, and auditable Ethereum SC vulnerability detection.
Eduardo Sardenberg Tavares, Antonio José G. Busson, Sérgio Colcher
Smart contracts are fundamental to blockchain ecosystems, but remain susceptible to security vulnerabilities that can lead to severe financial losses. Recent advances in agentic AI systems, powered by large language models (LLMs), enable autonomous code analysis and decision-making without explicit task-specific supervision. These systems leverage prompt engineering and zero-shot reasoning to detect vulnerabilities in smart contracts without prior fine-tuning. In this work, we evaluate the effectiveness of agentic LLM-based approaches in identifying vulnerabilities using prompt engineering and zero-shot reasoning across a curated dataset of Solidity smart contracts. Our findings highlight the limitations of current LLMs in automated vulnerability detection, providing insights into their practical applicability for securing decentralized applications. Our best-performing configuration, which integrates zero-shot reasoning with the Tree of Thoughts framework, achieved an F1-score of 73.66%.
Security vulnerabilities in smart contracts pose a serious threat to the blockchain ecosystem. Traditional detection methods heavily rely on expert-defined rules, leading to issues such as high false positives, false negatives, and poor scalability. Although recent deep learning-based approaches have shown promise, most focus on a single code modality, which limits their ability to capture the multi-dimensional characteristics of complex vulnerabilities. To address these limitations, this paper proposes MCLF, a novel smart contract vulnerability detection framework based on multimodal feature fusion and contrastive learning. The approach simultaneously extracts a Data Flow Graph and code semantic sequences from smart contracts to capture structural dependencies and contextual semantic information, respectively. A dual-branch Transformer architecture is designed to encode each type of feature, while contrastive learning is applied to optimize vector space alignment and enhance intra-class consistency. A masked language modeling task is further incorporated to improve semantic reasoning through multi-task joint training. Finally, transfer learning is employed to fine-tune the model for downstream vulnerability classification. Experimental results on a public benchmark dataset demonstrate that the proposed method achieves precision, recall, and F1score of 91.90%, 92.53% and 92.21%, respectively, significantly outperforming the most advanced vulnerability detection tools and validating the effectiveness of multimodal feature fusion and the contrastive learning mechanism.
Since the advent of smart contracts, security vulnerabilities have remained a persistent challenge, compromsing both the reliability of contract execution and the overall stability of the virtual currency market. Consequently... | Find, read and cite all the research you need on Tech Science Press
Rana Alaa, Darío González-Ferreiro, Carlos Beis-Penedo, Manuel Fernández‐Veiga · 6 authors
Split learning is an approach to collaborative learning in which a deep neural network is divided into two parts: client-side and server-side at a cut layer. The client side executes its model using its raw input data and sends the intermediate activation to the server side. This configuration architecture is very useful for enabling collaborative training when data or resources are separated between devices. However, split learning lacks the ability to verify the correctness and honesty of the computations that are performed and exchanged between the parties. To this purpose, this paper proposes a verifiable split learning framework that integrates a zk-SNARK proof to ensure correctness and verifiability. The zk-SNARK proof and verification are generated for both sides in forward propagation and backward propagation on the server side, guaranteeing verifiability on both sides. The verifiable split learning architecture is compared to a blockchain-enabled system for the same deep learning network, one that records updates but without generating the zero-knowledge proof. From the comparison, it can be deduced that applying the zk-SNARK test achieves verifiability and correctness, while blockchains are lightweight but unverifiable.
Fabian Hensel, Avik Banerjee, Elmira Ebrahimi, Stefan Schulte
Solidity smart contracts are widely used to implement decentralized applications. However, their development remains challenging due to the language’s domain-specific complexity, the immutability of deployed contracts, which prevents post-deployment fixes, and the high risk of introducing security-critical vulnerabilities. While Large Language Models (LLMs) have advanced code generation across general domains, they often struggle to meet the structural and security-specific demands of smart contract development. Therefore, this paper presents a domain-adapted code completion model trained on 22,000 labeled code constructs extracted from Solidity contracts. The model is built on a transformer-based architecture and fine-tuned using Quantized Low-Rank Adaptation (QLoRA), a parameter-efficient method. The dataset is processed to highlight secure coding patterns and structural semantics, enabling the model to learn from both preceding and succeeding contexts. Evaluation using perplexity, the Bilingual Evaluation Understudy (BLEU) score, and the Metric for Evaluation of Translation with Explicit Ordering (METEOR) shows significant improvements with consistent gains across all three metrics compared to the base model. These results demonstrate that targeted adaptation of language models can significantly enhance coding support in Solidity smart contracts.
Widespread adoption of Blockchain Technology has emphasized smart contracts as vital units of digital transactions. In the brief history of smart contracts, significant losses have occurred owing to unexplained vulnerabilities in blockchain-loaded contracts. Smart contract vulnerabilities endanger blockchain technology's viability and confidence. Blockchain-based applications rely on smart contracts to automate and trustlessly execute agreements. Their immutable and self-executing nature leaves them vulnerable to security vulnerabilities, which have caused major financial losses in decentralized platforms. Traditional static and symbolic analysis tools often miss sophisticated or obfuscated vulnerabilities, resulting in insufficient coverage or excessive false positives. These restrictions led the investigation of Deep Learning (DL) algorithms for detecting vulnerabilities in smart contract crucial since they can learn complicated patterns from code representations without manual feature engineering. Deep learning based methods for detecting vulnerabilities in smart contracts are thoroughly evaluated. We rigorously assess state-of-the-art DL models including BiLSTM, BiGRU, CNNLSTM, GCN, and CodeBERT using publicly available dataset. We provide a taxonomy of DL-based detection techniques and standardize performance criteria including accuracy, F1-score, and detection latency. Experimental results show model architecture trade-offs in detection, computational efficiency, and generalization.