The construction industry in developing countries continues to face significant challenges due to reliance on traditional, paper-based contract administration and management. This approach frequently results in inefficiencies, disputes, transparency issues and unethical practices. Although smart contracts enabled by blockchain technology present a promising solution to these longstanding issues, their adoption in developing countries remains limited. This study investigates the barriers to and strategies for the implementation of smart contracts within the construction industry, using Nigeria as a representative case. Adopting a qualitative research methodology, data were collected through semi-structured interviews with 14 experienced project managers selected via purposive sampling. A thematic analysis of the data identified several critical barriers, including resistance to change, low awareness, privacy concerns, legal uncertainties, technical hurdles, infrastructure deficits and economic instability. To overcome these barriers, the study proposes a strategic implementation framework informed by insights from interviewees and supported by literature. Key recommended strategies include educational and awareness initiatives, governmental support and policy reform, stakeholder collaboration, robust security measures, phased deployment and establishing supportive legal frameworks. The findings of this research offer valuable guidance for developing countries encountering similar constraints, providing a clear roadmap for successfully integrating smart contracts into construction practices.
Saad AL Azzam, Raenu Kolandaisamy, Ghassan AL Dharhani
Smart contracts (SCs) have become an essential component in the world of decentralized applications, automating transactions across blockchain networks without the need for intermediaries, and with this rise in adoption, the technology has also brought forth growing concern due to security vulnerabilities, which have led to serious financial damage, and the problem is far from being solved. Traditional auditing methods often struggle to capture the more intricate vulnerabilities hidden within smart contract logic, particularly owing to the irreversible nature of blockchain transactions. Given these challenges, researchers have been actively exploring more advanced detection techniques. Despite progress, many existing studies tend to focus narrowly on specific methods, whether static analysis, dynamic testing, or machine learning models, without offering a comprehensive comparison across all available approaches. This fragmented landscape leaves a noticeable gap for practitioners looking for a well-rounded understanding of smart contract security solutions. To address this, our study set out to systematically review the existing body of work, analysing 21 reviewed studies published between 2020 and 2024. The primary aim was to combine the diverse techniques that have been proposed for detecting vulnerabilities in smart contracts, ranging from static and dynamic analyses to more recent AI-driven models, graph-based techniques, and hybrid systems, critically evaluating their strengths, weaknesses, and practical effectiveness. The methodology followed a structured approach. We searched major research databases, IEEE Xplore, ACM Digital Library, SpringerLink, ScienceDirect, and Scopusโusing carefully crafted search queries to ensure that we captured the most relevant and up-to-date papers. Our findings revealed that AI-based methods, especially those leveraging deep neural networks and graph neural networks, have achieved impressive detection accuracy in controlled environments. For example, models such as ContractWard and SCVDIE-ENSEMBLE reported Micro-F1 scores of 98.48% and 95.46%, respectively, but these models also have a trade-offโthey demand high computational resources, which limits their real-world deployment in resource-constrained settings. On the other hand, lighter tools such as Slither and NeuCheck offer faster detection and lower resource usage but might fall short in regard to identifying more complex or new vulnerabilities. We also noticed a growing trend towards real-time monitoring tools, such as SODA and GPTScan, which aim to strike a balance by reducing false positives while providing proactive security measures. However, several challenges remain unresolved where many AI-driven models still rely heavily on labelled datasets, which may not generalize well to novel attack patterns. Scalability is another concern, especially for models that are computationally intensive.
ABSTRACT This paper proposes a solution for the issue of silent shareholders lacking influence over company decisions and not receiving adequate compensation. Thus, we adopt Palmon, Kleinman, and Medinetsโs (2022) โcapital contractโ framework and extend it by integrating smart contract functionality. This study then introduces a prototype to demonstrate how this enhanced framework can be implemented through blockchain-based smart contracts. By linking silent shareholdersโ dividends to executive compensation, these smart contracts enhance the trustworthiness and transparency of the compensation processes for executives and shareholders. What is more, blockchain-based smart contracts automate the contract terms, potentially reducing the need for intermediaries to monitor managerial actions. Also, smart contracts are flexible to meet diverse reporting requirements and adapt to the unique characteristics of a particular company. Data Availability: All data used in this study are available in the manuscript. JEL Classifications: M40; O33.
Amid the ongoing advancements associated with the Fourth Industrial Revolution and the intensification of digital transformation, the deployment of artificial intelligence (AI) within the banking sector has become an inevitable trajectory, enabling substantial innovations in financial management and operational processes. AI technologies facilitate the automation of complex workflows, reduce error rates, enhance operational efficiency, and improve customer experience through personalized services and accelerated response mechanisms. Applications span various functions, including customer onboarding, service delivery, product development, marketing, and risk management, thereby optimizing the banking value chain holistically. Moreover, AIโs capabilities in big data analytics and customer behavior prediction equip financial institutions with more robust decision-making tools that mitigate credit risk and fraud incidence. The convergence of AI and blockchain technologies further augments transaction security and transparency, thereby promoting the expansion of digital banking and decentralized finance ecosystems. This study aims to systematically examine the evolving roles and emerging applications of AI throughout the banking value chain, contributing to strategic frameworks oriented toward sustainable development within the digital era.
Do Tran Anh Duc, Lรช Thรกi Hรนng, Hoang-Phuong Chu-Nguyen, Van-Hau Pham ยท 5 authors
The increasing deployment of smart contracts has drawn significant attention to the urgent need for robust and scalable vulnerability detection techniques to mitigate substantial financial risks associated with their immutable nature on blockchain platforms. This paper introduces structured reasoning prompts using agent-role chaining for vulnerability detection that utilizes model capacity to enhance smart contract security through zero-shot and structured prompt engineering without fine-tuning. By carefully defining agent roles and embedding explicit reasoning steps within structured prompts for large language models (LLMs), the proposed method exploits the inherent reasoning capabilities of LLMs to identify security flaws in smart contracts without extensive model retraining. Experimental results demonstrate the effectiveness of the system in achieving competitive performance compared to existing vulnerability detection techniques, highlighting the potential of prompt engineering as an efficient and adaptable strategy for enhancing smart contract security.
Zhiyuan Wei, Jing Sun, Yuqiang Sun, Ye Liu ยท 13 authors
Blockchainโs inherent immutability, while transformative, creates critical security risks in smart contracts, where undetected vulnerabilities can result in irreversible financial losses. Current auditing tools and approaches often address specific vulnerability types, yet there is a need for a comprehensive solution that can detect a wide range of vulnerabilities with high accuracy. We propose LLM-SmartAudit, a novel framework that leverages Large Language Models (LLMs) to automate smart contract vulnerability detection and analysis. Using a multi-agent conversational architecture with a buffer-of-thought mechanism, LLM-SmartAudit maintains a dynamic record of insights generated throughout the audit process. This enables a collaborative system of specialized agents to iteratively refine their assessments, enhancing the accuracy and depth of vulnerability detection. To evaluate its effectiveness, LLM-SmartAudit was tested on three datasets: a benchmark for common vulnerabilities, a real-world project corpus, and a CVE dataset. It outperformed existing tools with 98% accuracy on common vulnerabilities and demonstrates higher accuracy in real-world scenarios. Additionally, it successfully identifies 12 out of 13 CVEs, surpassing other LLM-based methods. These results demonstrate the effectiveness of multi-agent collaboration in automated smart contract auditing, offering a scalable, adaptive, and highly efficient solution for blockchain security analysis.
This study utilizes version 6 of the regression analysis of time series (RATS) software package to implement the estimation of the bivariate diagonal generalized autoregressive conditional heteroscedasticity (GARCH) model combined with a composite asset selection approach including two hybrid performance measures to solve โthe trade-off problem between return and riskโ and โthe inconsistent results from different performance measuresโ in the problem of asset allocation within a group of minimum variance portfolios during the pre-COVID-19 and COVID-19 periods. Empirical results show that the optimal portfolios obtained from this approach and the assets added to a portfolio to achieve better performance differ between the pre-COVID-19 and COVID-19 periods. For instance, the optimal portfolios are the Chinese yuan-Ethereum and Bitcoin-Ethereum for the pre-COVID-19 period, but the WTI-Ethereum for the COVID-19 period. To achieve better performance, we added Ethereum to our portfolio during the pre-COVID-19 period, while WTI and Bitcoin were added during the COVID-19 period. Thus, the COVID-19 pandemic had a significant impact on the performance of asset allocation in the three markets. The proposed approaches in this study can be embedded in a computer as an asset allocation algorithm of Robo-advisers.
Smart contracts enable contract terms to be automatically executed and verified on the blockchain, and recent years have witnessed numerous applications of them in areas such as financial institutions and supply chains. The execution logic of a smart contract is closely related to the contract state, and thus the correct and safe execution of the contract depends heavily on the precise control and update of the contract state. However, the contract state update process can have issues. In particular, inconsistent state update issues can arise for reasons such as unsynchronized modifications. Inconsistent state update bugs have been exploited by attackers many times, but existing detection tools still have difficulty in effectively identifying them. This paper conducts the first large-scale empirical study about inconsistent state update vulnerabilities (that is, inconsistent state update bugs that are exploitable) in smart contracts, aiming to shed light for developers, researchers, tool builders, and language or library designers in order to avoid inconsistent state update vulnerabilities. We systematically investigate 116 inconsistent state update vulnerabilities in 352 real-world smart contract projects, summarizing their root causes, fix strategies, and exploitation methods. Our study provides 11 original and important findings, and we also give the implications of our findings. To illustrate the potential benefits of our research, we also develop a proof-of-concept checker based on one of our findings. The checker effectively detects issues in 64 popular GitHub projects, and 19 project owners have confirmed the detected issues at the time of writing. The result demonstrates the usefulness and importance of our findings for avoiding inconsistent state update vulnerabilities in smart contracts.
This paper explores the transformative impact of blockchain technology and smart contracts on the dynamics of trust within the financial sector. Trust is a cornerstone of financial transactions, traditionally established through centralized intermediaries and legal frameworks. However, the advent of blockchain technology introduces a decentralized, transparent, and tamper-resistant trust mechanism. This study aims to analyze how blockchain and smart contracts redefine financial trust by eliminating reliance on third-party intermediaries and automating trust through programmable agreements. Utilizing a mixed-methods approach, including case studies such as JP Morganโs Quorum blockchain platform, we examine the practical applications of these technologies and their effects on transactional efficiency, data privacy, and trust realization. Key findings reveal that blockchain significantly reduces transaction costs, enhances transparency, and increases security, paving the way for innovative financial products and services. The paper contributes to the understanding of how decentralized technologies are reshaping the future of financial trust and offers insights for regulators and financial institutions navigating this technological shift.
Yuchen Wang, Qiwen Wang, C.X. Ye, Shihong Zou ยท 5 authors
Off-chain execution schemes are a promising solution for enhancing block-chain scalability, reducing on-chain resource consumption, transaction costs, and processing time while improving data privacy. Nevertheless, off-chain execution schemes typically run smart contracts within small-scale execution groups. Since the number of nodes required to reach consensus is small, it is more likely that malicious nodes will gather in one execution group, causing the execution group to be controlled. We propose a reputation and risk-based off-chain group execution scheme for smart contracts. The scheme distributes potential malicious nodes across different execution groups, preventing any group from being dominated by malicious nodes. By integrating trust management mechanisms and a genetic algorithm, the proposed scheme mitigates the risk of malicious nodes colluding in a single group. This approach minimizes the likelihood of malicious node collusion and enhances the credibility of the execution group. Experiments demonstrate that the proposed scheme enhances blockchain throughput while increasing resilience against malicious attacks.
Yan Pang, Xiangfu Liu, Teng Huang, Yile Hong ยท 7 authors
Smart contract vulnerabilities have led to significant economic losses, threatening blockchain security and development. Graph neural network (GNN)-based approaches, which capture the structural properties of contracts and leverage code dependencies to better understand contract behavior, have become widely used for vulnerability detection. However, these approaches face challenges in losing valuable information during graph construction and failing to capture rich semantic content, while traditional GNNs struggle with long-range dependencies and global context in complex contract graphs. To address these challenges, we propose ConSense, a GNN-based Contract Sensing Framework for Smart Contract Vulnerability Detection. ConSense comprises two core components: the smart contract graph generator, which constructs contract graphs while retaining both structural and semantic information, and ExploreFormer, which effectively integrates local and global context using advanced attention mechanisms for vulnerability detection. Comprehensive experimental evaluations were performed on the IR-ESCD and SCVHunter-SCD datasets. For instance, the IR-ESCD benchmarkโwhich encompasses eight distinct vulnerability categoriesโdemonstrates that ConSense attains an average detection accuracy of 97.74%, with a mean processing time of 0.648 seconds per contract. These results signify a statistically significant improvement over state-of-the-art methods in both precision and computational efficiency.
The proxy design pattern allows Ethereum smart contracts to be simultaneously immutable and upgradeable, in which an original contract is split into a proxy contract containing the data storage and a logic contract containing the implementation logic. This architecture is known to have security issues, namely function collisions and storage collisions between the proxy and logic contracts, and has been exploited in real-world incidents to steal usersโ millions of dollars worth of digital assets. In response to this concern, several previous works have sought to identify proxy contracts in Ethereum and detect their collisions. However, they all fell short due to their limited coverage, often restricting analysis to only contracts with available source code or past transactions.To bridge this gap, we present Proxion, an automated cross-contract analyzer that identifies all proxy smart contracts and their collisions in Ethereum. What sets Proxion apart is its ability to analyze hidden smart contracts that lack both source code and past transactions. Equipped with various techniques to enhance efficiency and accuracy, Proxion outperforms the state-of-the-art tools, notably identifying millions more proxy contracts and thousands of unreported collisions. We apply Proxion to analyze over 36 million alive contracts from 2015 to 2023, revealing that 54.2% of them are proxy contracts, and about 1.5 million contracts exhibit at least one collision issue.
Love Opeyemi David, Marumo Kgomo, Clinton Aigbavboa
Introduction The traditional procurement system in the construction industry has been plagued by inefficiencies, often serving as a significant obstacle to project delivery. Thus, this study examines the dynamics of adopting smart contracts for project procurement for optimal project success and delivery, with insights and recommendations from the South African Construction Industry. Method The study employed a quantitative research approach utilizing descriptive and inferential statistics of Mean Item Score (MIS) and Exploratory Factor Analysis (EFA) for data analysis, based on a purposive sampling technique. Results The MIS results for the benefit, legal & regulatory constraints, and best practices of smart contracts range between 3.73 - 4.41 values, while the Kaiser-Meyer-Olkin (KMO) values were higher than the recommended 0.6 value for the EFA and Cronbach's Alpha value of 0.969 across the indicators. Discussion The study's findings revealed two categorized benefits of adopting smart contracts: administrative and operational efficiency of project procurement and procurement optimization; two components of legal and regulatory constraints: Transactional and legal encumbrance to smart contract implementation and legal gaps and ambiguity and two best practices: smart contract reliability practices for project procurement and consistent stakeholdersโ engagement for smart contract protocol standardization. The study concludes that Smart contracts can transform global project procurement within the construction industry. The study recommends the development of a green paper on smart contract adoption and integrating smart contracts into standard forms of construction contracts.
Akinde Michael Ogunmolu, Emonena Patrick Obrik-Uloho, Oluwaseun Oladeji Olaniyi, Aisha Temitope Arigbabu ยท 5 authors
This study investigates the systemic propagation of cyber risks between traditional financial institutions (TradFi) and decentralized finance (DeFi) infrastructures, focusing on oracles as critical conduits for contagion. Using publicly available datasetsโincluding MITRE ATT&CKยฎ for Financial Services, the Global Cybersecurity Index (GCI), and the REKT.news exploit archiveโthe study applies frequency analysis, logistic regression, time-series event studies, and Principal Component Analysis with cluster modeling to quantify institutional vulnerabilities, model breach likelihood, and evaluate governance impacts. Empirical findings show that API interconnectivity and DeFi exposure increase breach probabilities by up to 3.7 times, while countries in Cluster 0, such as Singapore and Estonia, exhibit governance indices 24โ28 points above average, correlating with lower systemic risks. Oracle-related incidents triggered over 150% volatility surges in TradFi-linked tokens like USDC and DAI, demonstrating oraclesโ role in cross-domain cyber risk transmission. The study recommends harmonizing cybersecurity governance frameworks across centralized and decentralized sectors, mandating periodic audits of oracle infrastructures, and developing integrated real-time threat monitoring systems to contain spillovers. These policy measures, alongside expanded cybersecurity workforce development, are essential to mitigate evolving cross-sector vulnerabilities. By combining rigorous empirical modeling with actionable recommendations, this research offers practical insights for policymakers, regulators, and cybersecurity professionals to strengthen resilience in the increasingly interconnected global financial ecosystem.
Smart contracts are a crucial component of blockchain systems, enabling high programmability and trusted transactions without the need for third parties. Their extensive implementation has significantly enhanced transparency and transactional efficiency in blockchain ecosystems. However, this advancement has also raised considerable concerns about vulnerabilities in smart contracts, which may lead to the hacking of blockchain applications, resulting in significant financial losses. This study suggests a traditional approach using Python regular expressions alongside the CodeBert deep learning model for smart contract vulnerability detection. Additionally, it compares the results of both approaches based on the number of instances. CodeBert emerges as an efficient deep learning model by detecting around 90.1 % of smart contract vulnerabilities.
This study aims to explore the integration of smart contracts within decentralized finance (DeFi) platforms, focusing on their security, functionality, and potential applications across various industries. Utilizing a comprehensive survey methodology, the research analyzes existing literature and case studies to identify the advantages and challenges associated with smart contracts. Key findings reveal that while smart contracts enhance efficiency and transparency, they also present significant security vulnerabilities that necessitate advanced mitigation strategies, including the use of artificial intelligence. The study concludes that addressing these challenges is crucial for the broader adoption of smart contracts, emphasizing the need for standardized frameworks and rigorous testing protocols. The implications of this research highlight the transformative potential of smart contracts in reshaping traditional business models and regulatory landscapes.
Based on the document content, I'll create a comprehensive abstract that captures the key aspects of the research. The research investigates the performance and efficiency of various consumer banking platforms using Grey Relational Analysis (GRA). The study analyzed five distinct banking platformsโOnline Banks (Nedbank's), Credit Unions, Peer-to-Peer (P2P) Lending, Fintech Solutions, and Cryptocurrency/Decentralized Finance (Deify)โacross four critical dimensions: Customer Satisfaction, Digital Banking and Technology, Financial Products and Services, and Customer Support. The analysis employed normalized data, deviation sequences, and grey relation coefficients to establish comprehensive performance metrics. The findings reveal significant variations in platform effectiveness, with Fintech solutions achieving the highest Grey Relationship Grade (GRG: 0.7387), followed closely by P2P lending (GRG: 0.7064). Traditional platforms like Credit Unions maintained moderate performance (GRG: 0.5674), while Cryptocurrency/Deify (GRG: 0.5117) and Online Banks (GRG: 0.5115) showed considerable room for improvement. The research demonstrates that success in modern banking requires a balanced integration of technological innovation with customer-centric services, rather than excellence in isolated areas. These results hold significant importance for shaping the strategic growth of banking services and guiding the future advancement of financial technology platforms.
This project represents a comprehensive digital transformation initiative designed to revolutionize traditional procurement practices through the development of an advanced Smart Tender Management System that enables vendors to seamlessly access complete tender documentation and specifications through a centralized online platform while facilitating efficient electronic bid submission processes. The system fundamentally addresses the inherent inefficiencies and cost burdens associated with conventional tendering methodologies by significantly minimizing additional operational expenses that traditionally encompass extensive advertising campaigns, physical document printing and distribution, manual handling procedures, and administrative overhead costs that often inflate the overall procurement budget. Through its sophisticated digital architecture, the application establishes stringent timeline management protocols that ensure the evaluation process adheres strictly to predetermined schedules and deadlines, thereby eliminating delays that frequently plague traditional tendering systems and compromise project timelines. The platform accommodates multiple vendor participation by providing a robust infrastructure that supports simultaneous bid submissions from diverse suppliers, contractors, and service providers, each presenting unique proposals with varying technical specifications, pricing structures, and implementation methodologies, from which procurement committees can systematically evaluate and select the most suitable proposals based on predetermined criteria including cost-effectiveness, technical merit, vendor credentials, and alignment with organizational objectives. This systematic approach to vendor selection and proposal evaluation has demonstrated significant potential for enhancing organizational profitability through optimized resource allocation, reduced procurement costs, improved vendor competition, and the selection of high-quality solutions that deliver superior value propositions. Furthermore, the implementation of this digital tendering system contributes substantially to improving the overall operational quality and efficiency of organizations by streamlining bureaucratic processes, reducing human error, enhancing transparency and accountability, facilitating better vendor relationships, and providing comprehensive audit trails that support compliance requirements and regulatory standards. The Smart Tender Management System's integration of advanced technologies, including secure document management, automated workflow processes, real-time communication capabilities, and comprehensive reporting mechanisms, positions it as a transformative solution that fundamentally reshapes how organizations approach procurement activities. In essence, this Smart Tender Management System represents a paradigmatic shift from traditional, paper-based, time-consuming procurement practices toward a modern, efficient, technology-driven approach that provides organizations with a powerful, comprehensive tool to systematically streamline their entire tendering ecosystem, significantly reduce operational and financial risks associated with procurement activities, enhance their competitive positioning in increasingly dynamic market environments, and establish sustainable procurement practices that support long-term organizational growth and success while maintaining the highest standards of transparency, efficiency, and stakeholder satisfaction throughout the entire tender lifecycle management process.
The integration of Artificial Intelligence (AI) into decentralized finance (DeFi) has triggered a paradigm shift in the automation and optimization of financial contracts, particularly within the domain of financial derivatives. Derivatives, including options, futures, swaps, and forwards, are among the most complex financial instruments, requiring accurate pricing, efficient settlement, and continuous risk monitoring. Smart contractsโself-executing agreements coded onto blockchain networksโhave emerged as a transformative mechanism to automate these processes. However, conventional smart contracts in DeFi are constrained by inefficiencies in execution logic, gas costs, vulnerability to adversarial trading strategies, and limitations in adapting to real-time market fluctuations. This manuscript investigates AI-driven optimization frameworks for smart contracts in derivatives markets, where machine learning algorithms, reinforcement learning agents, and predictive analytics are employed to dynamically enhance pricing mechanisms, counterparty risk management, and execution efficiency. The study builds on an extensive literature review of DeFi, AI-finance integration, and blockchain automation, proposing an AI-augmented smart contract architecture that enables adaptive fee structures, risk-adjusted margin calls, automated dispute resolution, and latency-sensitive derivatives clearing. A simulation-based methodology was employed, where deep reinforcement learning models interacted with synthetic market data to optimize contract logic in futures and options markets deployed on Ethereum Virtual Machine (EVM)-compatible blockchains. Statistical evaluation revealed that AI-enhanced smart contracts demonstrated 25โ40% improvement in transaction throughput, 18โ25% reduction in gas costs, 30โ35% enhancement in derivative pricing accuracy, and 50% reduction in settlement disputes compared to baseline blockchain contracts. The results highlight that AI-driven optimization is not only feasible but essential for scaling derivatives trading in DeFi to institutional-grade levels. The paper concludes by discussing regulatory implications, computational limitations, adversarial AI threats, and the future trajectory of autonomous financial engineering.
Non-fungible tokens (NFTs) have the potential to serve as fiduciary collateral in Indonesia. As a blockchain-based innovation, NFTs enable the unique representation and transfer of digital asset ownership. Under Indonesiaโs Fiduciary Security Law, NFTs qualify as fiduciary collateral objects since they are classified as intangible assets. This study examines copyright protection for NFTs in the context of fiduciary collateral, along with the legal and technical challenges in their implementation. Key obstacles include the lack of specific regulatory frameworks, insufficient blockchain infrastructure, and limited public understanding of NFTs as fiduciary collateral. Consequently, there is a need for comprehensive regulations and the establishment of oversight institutions to ensure transactional legality and security.Such regulatory measures are expected to facilitate the use of NFTs as fiduciary collateral, enhance public trust, and promote the growth of a blockchain-based digital ecosystem in Indonesia.
Jun 6, 2025ยทInnovations in Digital Finance and Intelligent Technologies: A Deep Dive into AI, Machine Learning, Cloud Computing, and Big Data in Transforming Global Payments and Financial Services
The financial sector is pervaded by high uncertainty and at a constant risk of damage to multiple stakeholders, either voluntarily or involuntarily. The highly unpredictable, multi-stakeholder, and multi-dimensional implications of machine learning assurances in finance have caused regulators around the world to impose strict regulations on their use. The heavy documentation requirements imposed on AI systems primarily aim to increase transparency through collaborative scrutiny of different stakeholders by allowing audits to be performed. This auditability requirement raises additional challenges for the implementation of distributed ledger-based systems and can discourage companies from utilizing the advantages such technologies convey. Nonetheless, operating in a system lacking collaborative transparency can pose even higher risks. Hence, the use of AI systems in finance needs to be adequately scrutinized in a manner that maintains the advantages of decentralization while ensuring the maintenance of internal and external compliance.
Smart contracts play a central role in automating processes on blockchain platforms, enabling operations such as identity management, access control, and logic execution. However, due to their inherent complexity and extensive interactions, they are often vulnerable to a range of security issues, such as reentrancy and timestamp dependency. In this work, we propose a novel vulnerability detection framework that integrates expert knowledge and multimodal representations to improve both accuracy and interpretability in smart contract analysis. Specifically, we leverage the Code Llama large language model to extract expert pattern features and construct three semantic graphs โ textual, opcode-level, and transaction-based โ to capture heterogeneous behavioral patterns of contracts. These multimodal features are aligned and fused through graph neural networks and attention mechanisms to form a unified representation, which is then used for binary classification of vulnerability presence. Experimental results on both public and custom datasets demonstrate that our approach achieves strong detection performance;for instance, the accuracy rate for detecting timestamp dependency vulnerabilities reaches 89 % on private datasets. These findings validate the effectiveness of combining multimodal fusion with large language model-guided knowledge extraction for enhanced smart contract security analysis.