Deep learning-based detectors have been widely proposed to predict vulnerabilities in smart contracts, yet their unreliable predictions pose severe security risks to financial transactions, making it critical to verify the reliability of vulnerability predictions. However, existing methods only produce prediction results, failing to provide an evidence chain to check whether these predicted vulnerabilities genuinely exist and deliver further guidance for fixing the vulnerabilities. Thus, making these vulnerability predictions verifiable remains an unexplored problem. In this paper, we propose SmartGuard, a novel verifiable vulnerability prediction framework for deep learning-based detectors and specifically designed for smart contracts. It integrates a deep learning-based detector with a symbolic prediction validator, where the latter acts as the backend formal engine to verify vulnerability predictions. Specifically, we present a graph-sequence multi-task learning model to detect vulnerabilities while generating transaction sequences that serve as evidence chains, explicitly revealing the triggering logic behind vulnerabilities. To bridge the gap between deep learning-based detectors and symbolic validators, we symbolically execute the generated transaction sequences against the verification conditions of vulnerability predictions. Furthermore, we propose a new metric, Vulnerability Prediction Suspiciousness (VPS), to evaluate the reliability of the predicted results. We implement SmartGuard on three representative types of vulnerabilities (Reentrancy, Ether-leaking, and Suicidal) to evaluate its performance in real-world scenarios. Our experimental results show that SmartGuard can effectively verify doubtful vulnerability predictions in real-world scenarios. It also outperforms state-of-the-art baselines by consistently reducing false reports by at least 15% across various Solidity versions. Case studies on complex contracts and DApps further demonstrate SmartGuard’s effectiveness in practice.
Smart contracts provide convenience to the financial industry by automating complex transactions without intermediaries. In the context of decentralized finance, ensuring the correctness, reliability, and efficiency of these automated systems is crucial, especially for loan management processes. This paper presents a formal specification of a smart contract-based loan management system using TLA+. The system is designed to handle various types of loans, including credit-based and collateral-backed loans, managing both regular and early repayments, as well as handling late payments and collateral forfeiture. Our specification defines key components such as loan states, repayment calculations, and collateral management, ensuring that all aspects of the loan lifecycle are rigorously modeled and verified. Safety properties, such as preventing negative loan balances and ensuring proper collateral ownership transfer, are enforced alongside liveness properties that guarantee the system reaches a termination state where all loans are either fully repaid or defaulted. We used the TLC model checker to verify the correctness of the system across all possible states and transitions. The verification process confirmed that the system consistently adheres to its formal specification under a variety of operational conditions. Through the formal specification and comprehensive verification with TLA+, this work enhances the dependability of smart contract based financial systems, providing a secure, verifiable, and resilient loan management framework ready for deployment in real-world DeFi environments.
Effective risk management solutions become absolutely crucial when financial markets embrace distributed technology and decentralized financing (DeFi). This study offers a thorough survey and comparative analysis of the integration of artificial intelligence (AI) in risk management for distributed arbitrage systems. We examine several modern caching techniques namely in memory caching, distributed caching, and proxy caching and their functions in enhancing performance in decentralized settings. Through literature review we examine the utilization of AI techniques for alleviating risks related to market volatility, liquidity challenges, operational failures, regulatory compliance, and security threats. This comparison research evaluates various case studies from prominent DeFi technologies, emphasizing critical performance metrics like latency reduction, load balancing, and system resilience. Additionally, we examine the problems and trade offs associated with these technologies, emphasizing their effects on consistency, scalability, and fault tolerance. By meticulously analyzing real world applications, specifically centering on the Aave platform as our principal case study, we illustrate how the purposeful amalgamation of AI with contemporary caching methodologies has revolutionized risk management in distributed arbitrage systems.
Few studies have investigated the delivery performance of suppliers in guarantor-intermediated trade finance with information delay in a multi-stage environment. This paper examines a model of dynamic guarantee finance (DGF) where guarantee institutions can adjust guarantee rates as an order progresses through different stages in the trade process, and compare it with the uniform guarantee financing model (the guarantee rate remains constant over the process, UGF). Additionally, we analyze the impact of information delay on the supplier's optimal delivery performance. The aim of this paper is to provide a foundational understanding of the potential benefits associated with transitioning from uniform-rate pricing to dynamic-rate pricing in guarantor-intermediated trade finance. The results indicate that DGF can incentivize suppliers to enhance their delivery performance, with the incentive value increasing as the trade process becomes lengthier. However, information delay has a negative effect on the incentive value of DGF. Smart contracts can complement DGF by accelerating the process of information verification. These findings provide guidance for the effective implementation of contract innovations (such as DGF) in guarantor-intermediated trade finance and highlight how the dynamic trade process and information delay impact the supplier's delivery performance.
Eashwar Sivakumar, Kiran Jot Singh, Paras Chawla, G. Geetha
The complex nature of smart contracts necessities the development of a novel adaptable framework. As blockchain technology continues to expand into diverse fields, the demand for secure, efficient, and transparent systems becomes increasingly critical. RBEDH is the integration of Role Based Access Control (RBAC), Event Driven Architecture (EDA) and Hybrid Functionalities. The framework is applied to the scholarly publishing sector, where academic integrity, transparency, and reliability are crucial for ensuring the credibility and trustworthiness of published research. The smart contracts are executed between authors, reviewers and publishers. Functionality test was conducted using Ganache, Ropsten, and Ethereum Mainnet. These evaluations confirmed the consistency and effectiveness of contract deployment. Vulnerabilities such as re-entrancy, integer overflow/underflow, unauthorized access were tested using Securify, Mythril, Smartcheck and Oynete. It is found that the system is secure and it not susceptible to any of the above vulnerability. Further timestamp dependency was tested through Manticore, Slither and Echidna and the test results indicate the absence of vulnerability. Scalability and performance analysis results proves that the proposed framework is better on the basis of average energy consumption, latency and memory requirement when compared with the existing literature.
Blockchain technology can generate decentralized consensus , thereby enhancing the verifiability and contractability of payoff-relevant states in a distribution channel or supply chain (e.g., seller cost and/or buyer value). Therefore, smart contracts can be written between supply chain partners such that transactions are automatically executed at state-dependent prices. The author examines the economic impacts of blockchain-enabled smart contracts on the equilibrium generation and allocation of surplus in a three-level supply chain. The model highlights the role of smart contracts in removing ex post asymmetric decision rights, which are otherwise inherently present under regular contracts where trade decisions are made based on constant transfer prices. As a result, the seller (under cost uncertainty) or the buyer (under value uncertainty) may hurt itself by signing a smart contract with the middleman, despite the improvement in the supply chain's total efficiency and surplus. Therefore, the equilibrium choice of a smart contract can be socially insufficient. Moreover, interestingly, a prisoner's dilemma may arise: The seller's and the buyer's dominant choice is the regular contract, whereas everyone would be better off if the smart contract is jointly adopted by all parties.
Rim Ben Fekih, Mariam Lahami, Salma Bradai, Mohamed Jmaïel
Defined as an agreement between multiple parties and systematically executed by a computer code, smart contracts enable trust-less execution without a third party. Despite the trusted implementations that smart contracts offer, including those based on standards, different security problems and vulnerabilities arise during their development and execution. To address these issues, multiple studies have proposed potential solutions, focusing particularly on the verification of smart contracts and considering the standard-based ones using formal verification techniques. However, the sheer amount of research makes it difficult to accurately articulate the state-of-the-art. To tackle this challenge, we propose a systematic literature review that deals with formal verification of ERC-based smart contracts. ERC (Ethereum Request for Comments) standards enable a range of functionalities, such as the creation and management of tokens. Thus, our review provides an overview of ERC standards and examines their related potential issues. Furthermore, we investigate existing solutions presented in 19 relevant studies published between 2019 and July 2023. We analyze and classify approaches to formal modeling, properties’ specification and techniques used in the verification of smart contracts. Finally, we discuss the research challenges and suggest some promising future directions to stir research efforts into this area.
Jie Cai, Jiachi Chen, Tao Zhang, Xiapu Luo · 6 authors
The reentrancy vulnerability is one of the most notorious vulnerabilities of smart contracts. It enables attackers to hijack the control flow of a smart contract by invoking a function as the entry point and then re-invoking a function as the reentry point before the execution of the entry point ends. Although several approaches have been proposed to detect this vulnerability, they still face two main limitations. Firstly, existing approaches oversimplify the rules for identifying entry and reentry points, and many even neglect reentry point identification during vulnerability detection. Secondly, most existing approaches overlook the flow of state variables that are not promptly updated, a critical aspect of the reentrancy vulnerability. To address the limitations mentioned above, this article proposes a novel static analysis framework for reentry vulnerability detection. We formulate the reentrancy vulnerability detection as entry and reentry point identification with the state variable flow tracking. Based on the insight that most smart contracts are implemented following various technical standards, we utilize static analysis with standard-based rules to identify potential entry and reentry points. This is achieved by detecting the presence of hijackable and exploitable operations inside the smart contract. Meanwhile, we also conduct state variable flow tracking by the static taint analysis. To verify the effectiveness of our proposed approach, we construct three different datasets. Then We compare our approach with eight state-of-the-art smart contract vulnerability detectors, and our tool outperforms these baselines in detecting more vulnerable samples with fewer false positive samples. Meanwhile, our approach achieves a relatively shorter detection time with better detection results, striking a trade-off between effectiveness and efficiency.
Michael Bresil, Pwc Prasad, Md Shohel Sayeed, Umar Ali Bukar
In the short history of smart contracts, substantial losses have occurred due to unaccounted vulnerabilities in the smart contracts loaded onto the blockchain. Vulnerabilities in smart contracts threaten the viability and confidence of blockchain technology. Machine and deep learning architectures have been increasingly proposed to assist with writing smart contracts and detecting vulnerabilities to minimize the risks of successful attacks. This paper presents the findings of deep learning vulnerability detection in smart contracts collated from selected research papers and provides an overview of the vulnerability detection architecture. This paper assessed each component of the overall architecture, individually and in combinations, to infer potential relationships in increasing detection rates in smart contract vulnerability detection tools. This study found that the area of detecting smart contract vulnerabilities is gravitating towards deep learning models, with a particular focus on combining neural networks in serial or parallel to achieve high detection results. Feature extraction of syntax and semantic information greatly determines the detection results of a model, which suggests a strong relationship between the use of source code and neural networks. Despite this relationship, given the inaccessibility and unavailability of smart contract source code, future work should focus on efficiently extracting features and context from opcodes, which are more readily available on the blockchain. Based on these findings, the study offers an in-depth discussion of the matters arising, open issues, and key aspects such as bytecode and opcode in smart contracts, failure of attention mechanisms, the rationale of hybrid models, dataset challenges, computational complexity in a large-scale blockchain environment, optimization strategies, Expunge and interdisciplinary approaches to smart contract vulnerability detection. These discussions enhance the practical relevance of the paper, highlighting the opportunities for scaling deep learning models for real-world blockchain applications.
Reentrancy vulnerabilities in Ethereum smart contracts have caused significant financial losses, prompting the creation of several automated reentrancy detectors.However, these detectors frequently yield a high rate of false positives due to coarse detection rules, often misclassifying contracts protected by anti-reentrancy patterns as vulnerable.Thus, there is a critical need for the development of specialized automated tools to assist these detectors in accurately identifying anti-reentrancy patterns.While existing code analysis techniques show promise for this specific task, they still face significant challenges in recognizing anti-reentrancy patterns.These challenges are primarily due to the complex and varied features of anti-reentrancy patterns, compounded by insufficient prior knowledge about these features.This paper introduces AutoAR, an automated recognition system designed to explore and identify prevalent anti-reentrancy patterns in Ethereum contracts.AutoAR utilizes a specialized graph representation, RentPDG, combined with a data filtration approach, to effectively capture anti-reentrancy-related semantics from a large pool of contracts.Based on RentPDGs extracted from these contracts, AutoAR employs a recognition model that integrates a graph auto-encoder with a clustering technique, specifically tailored for precise anti-reentrancy pattern identification.Experimental results show AutoAR can assist existing detectors in identifying 12 prevalent anti-reentrancy patterns with 89% accuracy, and when integrated into the detection workflow, it significantly reduces false positives by over 85%.* The corresponding authors.a reentrancy vulnerability in Ethereum, leading to substantial financial losses.
This study presents a comprehensive systematic review of Artificial Intelligence (AI) applications in DecentralizedFinance (DeFi), emphasizing AI’s pivotal role in mitigating the vulnerabilities and operational complexities inherentin permissionless financial systems. By systematically analyzing 39 peer-reviewed studies from major scholarlydatabases, the review identifies five dominant application domains: fraud detection, smart contract security, marketprediction, credit risk assessment, and decentralized governance. It examines the diverse range of AI methodsspanning machine learning, deep learning, graph neural networks, and reinforcement learning—and evaluates theircomparative performance and limitations. The findings reveal that AI not only enhances DeFi’s transparency, trust,and efficiency but also underpins emerging capabilities such as autonomous governance and adaptive marketmechanisms. Persistent challenges including data scarcity, cross-chain generalization, interpretability, andscalability—underscore the need for robust, explainable, and ethical AI solutions. The review concludes that AIconstitutes a foundational enabler for secure, transparent, and resilient decentralized financial ecosystems, andoutlines critical future research directions for integrating trustworthy intelligence into the evolving DeFi landscape. (PDF) A Systematic Review on the Application of Artificial Intelligence in Decentralized Finance. Available from: https://www.researchgate.net/publication/397514996_A_Systematic_Review_on_the_Application_of_Artificial_Intelligence_in_Decentralized_Finance [accessed Nov 11 2025].
Blockchain-based smart contracts have received a lot of attention recently. Smart contracts are increasingly being used in a variety of fields, and are touted for their ability to reduce costs, ensure contract fulfillment, and provide high security. This has led to the view that smart contracts can replace existing contracts. However, there is also a negative view that smart contracts are unlikely to replace existing contracts and will only be used in limited areas, citing the limitations of the code itself, the existence of bugs, and the possibility of hacking. These conflicting views reflect the fact that the smart contract debate is far from settled and many issues, such as the legal nature of smart contracts and the legal issues that have arisen or may arise from smart contracts, are still unclear. In this article, we will discuss the legal regulation of smart contracts, among other issues related to smart contracts. Recently, several countries, including some states in the U.S., have moved to incorporate smart contracts into their legal systems, which seems to be a way to provide a legal basis for the future use of smart contracts and the development of related technologies, as well as to regulate them through legislation. The need for legal regulation of smart contracts can be seen in the need to introduce the concept of cryptocurrencies or smart contracts, and to prepare for the possibility that transactions using smart contracts will be activated in the near future. The direction of legal regulation can be broadly categorized into enacting a basic law on smart contracts in the form of a single law, and amending individual laws related to smart contracts. This article opts for the enactment of a basic law on smart contracts in the form of a single law, and discusses factors to be considered when enacting a so-called basic law on smart contracts, as well as what a specific law should look like.
Asad Ul Islam Khan, Rasim Özcan, Mohamed Abbas Ibrahim
In this paper, we use the Empirical Bayes estimation and multiple linear regression approach to examine the impact of the top 5 cryptocurrencies’ crash risks on the G-7 and China equity markets’ crash risks. MATLAB was used to calculate the crash risks, while Stata software was employed for the econometric analysis. Three crash risk measures are usedto validate the robustness of the results: (i) the relative frequency of the number of crash days in the market, (ii) the monthly returns’ skewness, and (iii) the down-to-up volatility. Our findings indicate that overall crash risks of the top 5 cryptocurrencies are positively related with G-7 and Chinese stock markets’ crash risk. This suggests that the crash risk transmits from the crypto to the equity markets and the crashes in crypto can serve as a predictor in the stock markets. Furthermore, there is a negative correlation between the historical crash risks of the G-7 stock market and the present crash risks of the same stock market. This suggests that past stock market crashes can serve as a predictive factor for assessing the current risk of a stock market crash.
This paper explores the convergence of Blockchain and Internet of Things (IoT) technologies and their transformative potential in digital financial transactions. With the rise of decentralized systems, blockchain provides a secure, transparent, and immutable ledger for recording financial transactions, while IoT enables real-time data collection and automation in financial networks. The integration of these two technologies offers enhanced scalability, security, and efficiency, significantly improving the performance of digital financial systems. However, the combination of blockchain and IoT introduces new challenges, including security vulnerabilities, privacy concerns, and the need for effective mitigation strategies. The paper highlights key trends, such as the rise of smart contracts, tokenization, decentralized finance (DeFi) applications, and the incorporation of artificial intelligence (AI) and machine learning (ML) to optimize blockchain-IoT ecosystems. It also examines the evolving regulatory and ethical considerations for these hybrid systems. As these technologies continue to mature, understanding their potential and addressing their associated risks will be critical for the future of digital financial transactions.
Decentralized finance (DeFi) is a rapidly evolving blockchain technology that offers a new perspective on financial services through Web3 applications. DeFi offers developers the flexibility to create financial services using smart contracts, leading to a lack of standardized protocols and challenges in applying traditional finance models for risk assessment, especially in the early stages of adoption. The Maker protocol is a prominent DeFi platform known for its diverse functionalities, including loan services. This study focuses on analyzing the risk associated with Maker's loan portfolio by developing a risk model based on multiple Brownian motions and passage levels, with Brownian motions representing different collateral types and passage levels representing users' collateralization ratios. Through numerical experiments using artificial and real data, we evaluate the model's effectiveness in assessing risk within the loan portfolio. While our findings demonstrate the model's potential for assessing risk within a single DeFi project, it is important to acknowledge that the model's assumptions may not be fully applicable to real-world data. This research underscores the importance of developing project-specific risk assessment models for individual DeFi projects and encourages further exploration of other DeFi protocols.
Mihai Andronie, Roman Blažek, Mariana Iatagan, Renata Skýpalová · 16 authors
Research background: Big data-driven artificial Internet of Things (IoT) fintech algorithms can provide real-time personalized financial service access, strengthen risk management, and manage, monitor, and mitigate transaction operational risks by operational credit risk management, suspicious financial transaction abnormal pattern detection, and synthetic financial data-based fraud simulation. Blockchain technologies, automated financial planning and investment advice services, and risk scoring and fraud detection tools can be leveraged in financial trading forecasting and planning, cryptocurrency transactions, and financial workflow automation and fraud detection. Algorithmic trading and fraud detection tools, distributed ledger and cryptocurrency technologies, and ensemble learning and support vector machine algorithms are pivotal in predictive analytics-based risk mitigation, customer behavior and preference-based financial product and service personalization, and financial transaction and fraud detection automation. Credit scoring and risk management tools can offer financial personalized recommendations based on customer data, behavior, and preferences, in addition to transaction history, by generative adversarial and deep learning recurrent neural networks. Purpose of the article: We show that blockchain and edge computing technologies, generative artificial IoT-based fintech algorithms, and transaction monitoring and credit scoring tools can be harnessed in financial decision-making processes and loan default rate mitigation for transaction, payment, and credit process efficiency. Generative and predictive artificial intelligence (AI) algorithmic trading systems can drive coherent customer service operations, provide tailored financial and investment advice, and influence financial decision processing, while performing real-time risk assessment and financial and trading risk scenario simulation across fluctuating market conditions. Fraud and money laundering prevention tools, blockchain and financial transaction technologies, and federated and decentralized machine learning algorithms can articulate algorithmic profiling-based transaction data patterns and structures, credit assessment, loan repaying likelihood prediction, and interest rate and credit lending risk management by real-time financial pattern and economic forecast-based credit analysis across investment payment and transaction record infrastructures. Methods: Research published between 2023 and 2024 was identified and analyzed across ProQuest, Scopus, and the Web of Science databases by use of screening and quality assessment software systems such as Abstrackr, AMSTAR, AXIS, CADIMA, CASP, Catchii, DistillerSR, Eppi-Reviewer, MMAT, Nested Knowledge, PICO Portal, Rayyan, ROBIS, and SRDR+. Findings & value added: The main value added derived from the systematic literature review is that generative AI-based operational risk management, fraud detection, and transaction monitoring tools can provide personalized financial support and services and clarify financial and credit decisions and operations by financial decision-making process automation in dynamic business environments based on fraud detection capabilities and transaction data analysis and assessment. The benefits for theory and current state of the art are that credit risk and financial forecasting tools, artificial IoT-based fintech and generative AI algorithms, and algorithmic trading and distributed ledger technologies can be deployed in financial decision-making and customer behavior pattern optimization, credit score assessment, and money laundering and fraudulent payment detection. Policy implications reveal that investment management and algorithmic credit scoring tools can streamline financial activity operational efficiency, design financial planning analysis and forecasting, and carry out financial service and transaction data analysis for informed transaction decision-making and fraudulent behavior pattern and incident detection, taking into account credit history and risk evaluation and improving personalized experiences.