The rapid integration of artificial intelligence (AI) into high-stakes domains such as healthcare, finance, defense, and governance has created an urgent demand for transparent, auditable, and tamper-resistant decision-making frameworks. While AI models, particularly deep learning architectures, provide unparalleled predictive power, their opaque "black-box" nature often results in accountability gaps, regulatory non-compliance, and ethical challenges. Traditional logging mechanisms fail to capture the complexity and sensitivity of AI-driven decisions, especially in multi-stakeholder ecosystems. Blockchain technology, with its inherent features of immutability, decentralization, and verifiability, presents itself as a transformative solution to this problem. This manuscript proposes and evaluates blockchain-based logging systems for AI auditing, highlighting how distributed ledgers can establish immutable trails of model inputs, intermediate reasoning, and final outputs. The study conducts a comprehensive literature review on AI auditability, trust mechanisms, and blockchain applications, followed by a methodological framework integrating permissioned blockchains with explainable AI (XAI). A statistical analysis is presented to compare blockchain-logging versus traditional logging systems in terms of latency, transparency, energy consumption, scalability, and regulatory compliance. Results indicate that blockchain-based logging improves transparency by 78%, strengthens compliance traceability by 65%, and reduces auditing disputes by 52%, albeit at a moderate computational cost. The paper concludes that blockchain-based logging is not merely a technical enhancement but a regulatory and ethical necessity for next-generation AI systems. Future research directions include hybrid blockchain models, privacy-preserving logging protocols, and AI-governed adaptive consensus mechanisms.
Blockchain networks securely record transactions and enable decentralised transactions using cryptocurrencies. However, the pseudonymity nature of the participants makes the blockchain network a platform for illegal activities, such as money laundering, which poses significant threats to financial security and regulatory compliance. Money laundering activities undermine the integrity of financial systems, foster criminal enterprises, and enable tax evasion. This research explores the impact of timestamp-based (Time step) features in detecting money laundering activities within the Bitcoin network, utilising the Elliptic++ dataset. A correlation-based analysis revealed that the first block appeared in feature was the most strongly correlated with the Time step. Additionally, classification results highlighted XGBoost as the most effective classifier, with the first block appeared in feature identified as the most influential, based on Shapley values from the eXplainable Artificial Intelligence (XAI) technique.
R Sheeba, Jay Prakash Mahto, Syed Sabith Ansari, Zian Rajeshkumar Surani · 6 authors
The model presented in this work represents a paradigm shift that sets a completely novel standard for data distributed validation in ethical AI training. Our new paradigm integrates fault-tolerant Byzantine consensus along with zero-knowledge proofs for secured and provable auditing of data within decentralized AI systems. The framework uses a two-layer blockchain design that separates metadata anchoring from validation logs, allowing it to achieve an instantaneous compliance check time of less than one second while maintaining privacy compliance to GDPR. Key innovations comprise a sharded Merkle-Patricia Trie kind for dynamic data lineage chains, the application of differential privacy and federated learning with bias-neutralizing validation oracles, as well as the design of incentive engineering under a non-Markovian reward system for multiple agents. The results of experimentation prove that, under adversarial conditions, the detection of anomalies is 40% faster than centralized alternatives, while maintaining an integrity verification of the audit trail at 99.99%. The collaboration between AI explainability matrices and post-quantum secure voting mechanisms in this work set innovative standards for decentralized ethical oversight of mission-critical operations, thus transforming the trust dynamics among model developers, data subjects, and auditors.
Alexandra Vultureanu‐Albişi, Costin Bădică, Mirjana Ivanović
The Internet of Things (IoT) paradigm is evolving and the Next-Generation IoT (NG-IoT) ecosystem will incorporate distributed ledger and blockchain technology, AI-adapted components, and intelligent edge solutions that take advantage of edge computing, Artificial Intelligence (AI), networks, and communications. In addition to the low integration of eXplainable Artificial Intelligence (XAI) in the IoT or NG-IoT contexts, the explainability of these systems is rarely evaluated. Due to these limitations, we thoroughly examined the current state of XAI integration with IoT services. We propose a new conceptual framework called eXING-IoT (eXplainability Integrated in the Next Generation IoT) for better NG-IoT systems' explainability integration and evaluation. This includes a list of qualities that future NG-IoT environments should have, thus paving the way for the advancement of NG-IoT beyond the state of the art.
Open access
Explainable Artificial Intelligence (XAI)
Scientific Computing and Data Management
Artificial Intelligence in Healthcare and Education
With the rise of machine learning techniques, ensuring the fairness of decisions made by machine learning algorithms has become of great importance in critical applications. However, measuring fairness often requires full access to the model parameters, which compromises the confidentiality of the models. In this paper, we propose a solution using zero-knowledge proofs, which allows the model owner to convince the public that a machine learning model is fair while preserving the secrecy of the model. To circumvent the efficiency barrier of naively proving machine learning inferences in zero-knowledge, our key innovation is a new approach to measure fairness only with model parameters and some aggregated information of the input, but not on any specific dataset. To achieve this goal, we derive new bounds for the fairness of logistic regression and deep neural network models that are tighter and better reflecting the fairness compared to prior work. Moreover, we develop efficient zero-knowledge proof protocols for common computations involved in measuring fairness, including the spectral norm of matrices, maximum, absolute value, and fixed-point arithmetic. We have fully implemented our system, FairZK, that proves machine learning fairness in zero-knowledge. Experimental results show that FairZK is significantly faster than the naive approach and an existing scheme that use zero-knowledge inferences as a subroutine. The prover time is improved by 3.1x--1789x depending on the size of the model and the dataset. FairZK can scale to a large model with 47 million parameters for the first time, and generates a proof for its fairness in 343 seconds. This is estimated to be 4 orders of magnitude faster than existing schemes, which only scale to small models with hundreds to thousands of parameters.
Samreen Khan, Suguna Balusamy, Mariya Princy Antony Saviour, Satish Bojjawar · 6 authors
Blockchain Technology (BT) is a promising approach for building scalable & reliable open Decision Support Systems and Artificial Intelligence (AI) could be integrated with BT to present a better approach towards the same. Transparency, trust, data security, and scalability are traditionally regarded as the major deficiencies of DSS models which may impede their efficacy in critical decision-making environments. Seamlessly integrating blockchain with AI-powered DSS can help with data integrity, as well as prevent manipulation and inconsistencies while improving trust with many different stakeholders, because all transactions on a blockchain are transparent immutable, owing to blockchain's decentralized ledger system. The paper investigates a hybrid Blockchain-AI framework that can overcome the main drawbacks of existing DSS systems. This means that AI predictive analytics and machine learning models can match massive datasets for decision making, while blockchain guarantees the integrity, traceability, and decentralization of the data's inputs and outputs. Smart contracts: automate decision processes & ensure tamper-proof execution for defined set of rules & policies In addition, blockchain consensus mechanisms (e.g., Proof-of-Stake, Byzantine Fault Tolerance) can facilitate transparency and verifiability, thus improving conflict-avoiding distributed decision-making. To enable privacy-preserving collaboration between organizations, we introduce a new Decentralized AI Decision Support System (Dai-DSS) framework that integrates federated learning and a distributed ledger. It enables real-time data sharing while preserving data sovereignty, which helps organizations comply with regulatory standards such as GDRP and HIPAA. We are also looking at various optimization techniques such as off-chain scaling solutions (sidechains, Layer-2 protocols, etc.) that improve system efficiency while still keeping the system decentralized. Our research shows that by comparing with the traditional DSS models, the Blockchain-AI integration strengthens system resilience, improves the elimination of single point failure and facilitates open, trustless decision-making processes. Practical applications and the benefits of the proposed model are demonstrated through case studies in healthcare, finance, and supply chain management. Lastly, we address the challenges including computational overhead, interoperability and regulatory challenges, and suggest future research avenues on AI-augmented consensus algorithms and quantum-resistant cryptography. The research demonstrates the ability of Blockchain-AI synergy to revolutionize decision-making systems through decentralized, clean, and scalable operating frameworks across industries.
Verification of the integrity of deep learning inference is crucial for understanding whether a model is being applied correctly. However, such verification typically requires access to model weights and (potentially sensitive or private) training data. So-called Zero-knowledge Succinct Non-Interactive Arguments of Knowledge (ZK-SNARKs) would appear to provide the capability to verify model inference without access to such sensitive data. However, applying ZK-SNARKs to modern neural networks, such as transformers and large vision models, introduces significant computational overhead. We present TeleSparse, a ZK-friendly post-processing mechanisms to produce practical solutions to this problem. TeleSparse tackles two fundamental challenges inherent in applying ZK-SNARKs to modern neural networks: (1) Reducing circuit constraints: Over-parameterized models result in numerous constraints for ZK-SNARK verification, driving up memory and proof generation costs. We address this by applying sparsification to neural network models, enhancing proof efficiency without compromising accuracy or security. (2) Minimizing the size of lookup tables required for non-linear functions, by optimizing activation ranges through neural teleportation, a novel adaptation for narrowing activation functions' range. TeleSparse reduces prover memory usage by 67% and proof generation time by 46% on the same model, with an accuracy trade-off of approximately 1%. We implement our framework using the Halo2 proving system and demonstrate its effectiveness across multiple architectures (Vision-transformer, ResNet, MobileNet) and datasets (ImageNet,CIFAR-10,CIFAR-100). This work opens new directions for ZK-friendly model design, moving toward scalable, resource-efficient verifiable deep learning.
Muhammad Kamran, Maaz Rehan, Muhammad Maaz Rehan, Wasif Nisar · 6 authors
Blockchain technology is increasingly being adopted across critical domains, such as healthcare and finance, yet it remains susceptible to anomalies and malicious attacks. Hence, robust anomaly detection is essential in these decentralized systems to maintain integrity, trust, and reliability. However, anomaly detection is still challenging due to data imbalances, adversarial resilience, and the lack of explanation in existing approaches. This work presents ARCADE, a novel approach for adversarially resilient anomaly detection in blockchain networks that leverages an optimized cost-sensitive stacking ensemble learning combined with explainable artificial intelligence (XAI) techniques. Firstly, the proposed approach uses cost-sensitive learning to address the data imbalance problem by optimizing class weights that are integrated with stacking ensemble learning to enhance detection accuracy. Secondly, along with this, newly engineered features are employed to strengthen the resilience of the model against malicious perturbations. Lastly, XAI techniques are applied to provide comprehensive insights and explanations for model prediction. To evaluate ARCADE, the Ethereum network transactions dataset is utilized to ensure a realistic case study. The experimental results show the superiority of the ARCADE in several aspects, achieving a high accuracy of 99.65%; strong resilience against adversarial perturbations, achieving an accuracy of 99.38% for low-intensity attacks, 91.04% for moderate attacks, and over 78% for extreme attacks; and surpassing existing techniques while also providing explainability for domain users.
The decentralized, open-source architecture of blockchain technology, exemplified by the Ethereum platform, has transformed online transactions by enabling secure and transparent exchanges. However, this architecture also exposes the network to various security threats that cyber attackers can exploit. Detecting suspicious behaviors in account on the Ethereum blockchain can help mitigate attacks, including phishing, Ponzi schemes, eclipse attacks, Sybil attacks, and distributed denial of service (DDoS) incidents. The proposed system introduces an ensemble stacking model combining Random Forest (RF), eXtreme Gradient Boosting (XGBoost), and a neural network (NN) to detect potential threats within the Ethereum platform. The ensemble model is fine-tuned using Bayesian optimization to enhance predictive accuracy, while explainable artificial intelligence (XAI) tools-SHAP, LIME, and ELI5-provide interpretable feature insights, improving transparency in model predictions. The dataset used comprises 9,841 Ethereum transactions across 52 initial fields (reduced to 17 relevant features), encompassing both legitimate and fraudulent records. The experimental findings demonstrate that the proposed model achieves a superior accuracy of 99.6%, outperforming that of other cutting-edge methods. These findings demonstrate that the XAI-enabled ensemble stacking model offers a highly effective, interpretable solution for blockchain security, strengthening trust and reliability within the Ethereum ecosystem.
In the modern era of digitalization, integration with blockchain and machine learning (ML) technologies is most important for improving applications in healthcare management and secure prediction analysis of health data. This research aims to develop a novel methodology for securely storing patient medical data and analyzing it for PCOS prediction. The main goals are to leverage Hyperledger Fabric for immutable, private data and to integrate Explainable Artificial Intelligence (XAI) techniques to enhance transparency in decision-making. The innovation of this study is the unique integration of blockchain technology with ML and XAI, solving critical issues of data security and model interpretability in healthcare. With the Caliper tool, the Hyperledger Fabric blockchain's performance is evaluated and enhanced. The suggested Explainable AI-based blockchain system for Polycystic Ovary Syndrome detection (EAIBS-PCOS) system demonstrates outstanding performance and records 98% accuracy, 100% precision, 98.04% recall, and a resultant F1-score of 99.01%. Such quantitative measures ensure the success of the proposed methodology in delivering dependable and intelligible predictions for PCOS diagnosis, therefore making a great addition to the literature while serving as a solid solution for healthcare applications in the near future.
Open access
Impact of AI and Big Data on Business and Society
FinTech, Crowdfunding, Digital Finance
Artificial Intelligence in Healthcare and Education
Abstract As machine learning technologies advance rapidly across various domains, concerns over data privacy and model security have grown significantly. These challenges are particularly pronounced when models are trained and deployed on cloud platforms or third-party servers due to the computational resource limitations of users’ end devices. In response, zero-knowledge proof (ZKP) technology has emerged as a promising solution, enabling effective validation of model performance and authenticity in both training and inference processes without disclosing sensitive data. Thus, ZKP ensures the verifiability and security of machine learning models, making it a valuable tool for privacy-preserving AI. Although some research has explored the verifiable machine learning solutions that exploit ZKP, a comprehensive survey and summary of these efforts remains absent. This survey paper aims to bridge this gap by reviewing and analyzing all the existing Zero-Knowledge Machine Learning (ZKML) research from June 2017 to August 2025. We begin by introducing the concept of ZKML and outlining its ZKP algorithmic setups under three key categories: verifiable training, verifiable inference, and verifiable testing. Next, we provide a comprehensive categorization of existing ZKML research within these categories and analyze the works in detail. Furthermore, we explore the implementation challenges faced in this field and discuss the improvement works to address these obstacles. Additionally, we highlight several commercial applications of ZKML technology. Finally, we propose promising directions for future advancements in this domain.
This talk given at the 2025 MDIC CM&S Summit on "Credible Models in the AI Age" traces the evolution of trust mechanisms in computational systems, from traditional model validation approaches in mechanistic modeling to emerging cryptographic verification methods for AI. We'll explore how the credibility challenge for regulators has transformed as we've moved from deterministic simulations to probabilistic AI systems, and examine how cryptographic proofs, zero-knowledge techniques, and verifiable computation are creating new pathways for establishing trust in AI outputs. By understanding this historical progression, we can better appreciate both the continuity and fundamental shifts in how we ensure reliability in our computational approaches.
Naresh Kumar Satish, Mathieu Mercadier, Cristina Hava Muntean, Anderson Augusto Simiscuka
The cryptocurrency market is widely regarded as one of the most volatile financial markets due to inconsistencies in its pricing factors. Despite this volatility, it continues to attract a large population of investors, many of whom incur significant losses. To address this challenge and support risk assessment for investors, users, and other stakeholders, this paper focuses on forecasting Ethereum prices by analyzing social media sentiment. The study gathers data from sources such as global news headlines and Reddit discussion forums, enhancing it with hybrid sentiment features derived from the VADER, BERT and TextBlob models. These sentiment insights are then correlated with Ethereums financial parameters to establish meaningful relationships within the data, which are used to train machine learning models. The study evaluates the predictive performance of Random Forest, Extreme Gradient Boosting, and Long Short-Term Memory models. Among these, Extreme Gradient Boosting demonstrated superior performance, effectively capturing complex relationships within the data and achieving an R-squared value of 0.982115. To further enhance the studys risk assessment capabilities, the concept of Explainable Artificial Intelligence (XAI) is employed to improve transparency and accountability in the model outcomes. Specifically, Shapley Additive Explanations (SHAP) are used to interpret the feature interactions within the Extreme Gradient Boosting model, thereby increasing its reliability and providing deeper insights into its decision-making process.
Abstraction Liquidity Theory (ALT) develops a formal framework for determining when local problem-solving traces become reusable abstraction assets that reduce downstream search, evaluation, and certification costs. The paper treats abstractions as operational tokens rather than informal artifacts, and evaluates them through declared receivers, opportunity measures, baselines, lifecycle costs, telemetry, evidence validity, transport scope, authority envelopes, hazard constraints, and runtime certificate packets. The manuscript introduces an actor-neutral certification kernel for AI agents and other computational actors. It specifies machine-readable packet schemas, dual exploration and settlement ledgers, finite-sample lower and upper bounds, causal and calibrated-proxy value estimands, mission-validity certificates, adversarial-token rejection, root/finality checks, baseline refresh, deprecation, resurrection, rollback, and kernel-update bridges. The goal is to make abstraction evaluation executable: an agent should be able to parse a packet, verify evidence, admit or reject a token, suspend stale claims, deprecate negative-liquidity tokens, and preserve raw net safe capital under fail-closed rules. The paper further defines Target-valid ALT-CARA, a criterion for certified ASI realization acceleration. Rather than claiming unconstrained ASI achievement, ALT-CARA formalizes time-to-target acceleration relative to a resource-matched baseline upper envelope, under declared capability bases, target-validity certificates, raw net solvency, viability conditions, hazard and authority constraints, transport validity, finality, and causal reproduction evidence. The framework connects AI evaluation, causal inference, runtime verification, risk control, skill reuse, safe exploration, and distributed certification into a single theory of mission-valid safe abstraction capital.
This study investigates the critical challenges associated with ensuring the security and robustness of artificial intelligence (AI) systems, especially within high-stakes applications such as autonomous vehicles, healthcare, and financial technologies. The primary objective is to identify vulnerabilities in AI algorithms and propose effective mitigation strategies. The research emphasizes contemporary threats, including adversarial attacks, algorithmic opacity, data breaches, and the ethical ramifications of AI deployment. A review of current literature reveals that adversarial attacks, where subtle input perturbations cause significant misclassifications, present a considerable risk to AI reliability. Techniques such as robust training, involving training models on adversarial examples, have shown effectiveness in improving resilience, albeit with higher computational demands. The study also explores the importance of explainable AI (XAI) tools like LIME and SHAP, which enhance transparency by clarifying the decision-making processes of complex models. This transparency is vital for fostering user trust, especially in fields like medicine and finance, where understanding AI decisions is essential. XAI approaches enable better oversight and adherence to ethical standards. Data privacy concerns are addressed through methods such as differential privacy, which protects sensitive information by adding noise, and federated learning, which enables decentralized model training without exposing raw data. The findings indicate that these strategies secure data while maintaining model efficacy. By integrating robustness and explainability, this study contributes practical solutions to strengthen AI systems against evolving threats, advancing AI security and fostering trust in these technologies.
Blockchain smart contracts have emerged as a transformative technology, enabling the automation and execution of contractual agreements. These self-executing software programs leverage blockchain’s distributed and immutable nature to eliminate the need for third-party intermediaries. However, this new paradigm of automation and authority introduces a complex environment with technical intricacies that users are expected to understand and trust. The irreversible nature of blockchain decisions exacerbates these issues, as any mistake or misuse cannot be rectified. Current smart contract designs often neglect human-centric approaches and the exploration of trustworthiness characteristics, such as explainability. Explainability, a renowned requirement in Explainable Artificial Intelligence (XAI) aimed at enhancing human understandability, transparency and trust, has yet to be thoroughly examined in the context of smart contracts. A noticeable gap exists in the literature concerning the early development of explainability requirements, including established methods and frameworks for addressing requirements analysis phases, design principles, evaluation of their necessity and trade-offs. Therefore, this thesis aims to advance the field of blockchain smart contract systems by introducing explainability as a design concern, fundamentally prompting requirements engineers and designers to cater to this concern during the early development phases. Specifically, we provide guidelines for explainability requirements analysis, addressing what, why, when and to whom to explain. We propose design principles for integrating explainability into the early stages of development. To tailor explainability further, we propose a human-centred framework for determining information requirements in smart contract explanations, utilising situational awareness theories to address the ‘what to explain’ aspect. Additionally, we present ‘explainability purposes’ as an integral resource in evaluating and designing explainability. Our approach includes a novel evaluation framework inspired by the metacognitive explanation-based theory of surprise, addressing the ‘why to explain’ aspect. The proposed approaches have been evaluated through qualitative validations and expert feedback. We have illustrated the added value and constraints of explainability requirements in smart contracts by presenting case studies drawn from literature, industry scenarios and real-world projects. This study informs requirements engineers and designers regarding how to elicit, design and evaluate the need for explainability requirements, contributing to the advancement of the early development of smart contracts.
Artificial intelligence (AI) has both capabilities and limitations in managing uncertainty, particularly in handling average cases versus extreme outliers. AI excels at analyzing large datasets and predicting typical outcomes, but struggles with rare, critical scenarios that require flexibility beyond its data-driven approach. Integrating human expertise with AI, especially in managing anomalies, enhances AI's potential to address complex situations. The discussion also highlights the tension between rapid innovation and rigid governance, emphasizing the importance of trust and adaptable frameworks for progress in technology and finance. Governance must evolve to allow faster, individualized solutions while maintaining oversight to prevent risks. The concept of "digital twins" improves adaptability and cost efficiency by modeling and simulating physical entities. Transitioning from hardware- to software-driven solutions underscores the need for production agility, enabling industries to adapt without major structural changes. The analysis addresses AI's role in large-scale societal issues like decarbonization, stressing the importance of managing not just mean outcomes but also catastrophic tail risks. AI must identify gaps in its understanding, enhancing efficiency through collaborative training. Technological advancements, including sensors and digital twins, enhance AI's real-time analysis capabilities, expanding applications in robotics, precision agriculture, and decentralized healthcare. The interplay between AI, governance, and innovation shows potential for solving major challenges in finance, healthcare, and sustainability, contingent on integrating adaptive human-AI collaboration.
Blockchain technology has gained popularity due to its key features of decentralization, cryptographic verification, and immutability, which have proven extremely useful in various industries. However, despite their impressive security features, blockchain networks are not immune to cyber threats. In recent times, the blockchain system has been threatened by fraudulent attacks that require quick responses. Machine learning and deep learning models are increasingly leveraged to address these challenges. However, due to their black box nature, these models lack transparency, which is a major criticism. This study presents an approach to enhancing fraud detection mechanisms on Ethereum. This study presents an efficient and transparent fraud detection system on Ethereum known as IFS-TABPFN. An interpretable feature selection approach based on Shap values and optimized gradient boosting was introduced to develop five deep learning models built on neural networks. These models included Multilayer Perceptron (MLP), Long Short-Term Memory (LSTM), Convolutional Neural Networks (CNN), Convolutional Neural Networks and Long Short-Term Memory (CLSTM) and Tabular Prior-Data Fitted Network (TabPFN). A comparative analysis of our results indicates that IFS-TABPFN achieves 99.2% accuracy in just a few seconds, outperforming other neural networks and existing systems. This study highlights the importance of explainable AI in understanding how features influence decisions, performance and contribute to artificial intelligence models' transparency and trust.
Paolo Giudici, Alessandro Piergallini, Maria Cristina Recchioni, Emanuela Raffinetti
We consider the problem of developing explainable Artificial Intelligence methods to interpret the results of Artificial Intelligence models for time series data, taking time dependency into account. To this end, we extend the Shapley–Lorenz method, normalised by construction, to Artificial Intelligence for time series, such as neural networks and recurrent neural networks. We illustrate the application of our proposal to a time series of Bitcoin prices, which acts as the response variable, along with time series of classical financial prices, which act as explanatory variables. Three main findings emerge from the analysis. First, recurrent neural networks lead to a better performance, in terms of accuracy and robustness, with respect to classic neural networks. Second, the best performing models indicate that Bitcoin prices are affected mostly by their lagged values, and that their explainability, in terms of classical financial assets, is limited. Third, although limited, the contribution of classical assets to Bitcoin price prediction is well captured by recurrent neural networks.