Purpose This study aims to qualitatively investigate and uncover the factors influencing resistance intention to local digital wallets in the context of Brunei Darussalam. Design/methodology/approach This study uses a qualitative approach to investigate resistance toward adopting local digital wallets in Brunei Darussalam. Using a purposive sampling approach, an online survey was distributed via prevalent social media platforms, targeting individuals who intentionally refrain from using these services. Structured questionnaires capture demographic data, while open-ended questions explore reasons for resistance. Thematic analysis, guided by the innovation resistance theory (IRT), extracts and analyses collected responses to reveal emergent themes. Data saturation, anticipated with a minimum of 12 responses, signifies sample adequacy. Findings This study collects 55 responses reflecting resistance to local digital wallet adoption, primarily from females and youths (i.e. Generation Z groups). Through deductive thematic analysis, guided by the IRT framework, five emerging themes related to resistance behavior were identified. The “Access Barrier” highlights issues such as the lack of prerequisites like bank accounts, limited capital and restricted internet access; the “Risk Barrier” emphasizes concerns about data security, privacy, financial fraud and money management; the “Tradition Barrier” underscores challenges associated with preferences for cash and plastic money; the “Usage Barrier” addresses issues of digital overload, process complexity and unfamiliarity; and finally, the “Value Barrier” illuminates issues of apathy, lifestyle incompatibility and limited vendor acceptance. Research limitations/implications This study recognizes several limitations related to sample representativeness, reliance on self-reported data, the use of a cross-sectional design and a narrow focus on individual perspectives. Moving forward, it is important for future studies to address these limitations by implementing strategies such as obtaining more balanced samples, incorporating quantitative data collection methods, adopting longitudinal approaches and incorporating perspectives from a wider range of stakeholders. By doing so, researchers can enhance the validity and generalizability of their findings, leading to a more robust understanding of the dynamics surrounding digital wallet resistance and adoption within academic discourse. Practical implications Digital wallet providers must prioritize accessibility by addressing barriers like prerequisites, financial constraints and internet connectivity. Strategies include expanding access to unbanked populations, incentivizing first-time users and enhancing internet infrastructure. To mitigate risk-related barriers, providers should focus on data security, privacy protection and user safety. Tailoring offerings to accommodate diverse user preferences, simplifying usage processes, increasing awareness through educational campaigns and collaborating with stakeholders are essential. Collaboration between providers, regulatory bodies, financial institutions and merchants is vital to address multifaceted barriers, improve infrastructure, enhance security measures and promote user awareness, enabling digital wallet adoption and fostering the growth of the digital economy. Originality/value This study contributes to understanding the factors influencing the resistance behaviour to wards local digital wallets in Brunei Darussalam, offering insights for policymakers, academicians and digital wallet providers to address resistance and promote sustainable adoption.

Graph anomaly detection (GAD) refers to identifying abnormal graph nodes or edges that heavily deviate from normal observations. Existing approaches inevitably suffer from the influence of imbalanced data and privacy protection. This shortcoming poses challenges in optimizing node embeddings and detecting multitype anomalies simultaneously, resulting in decreased accuracy of existing GAD models. To address this shortcoming, we introduce a new federated learning model for graph anomaly detection (FedGAD). FedGAD enables collaborative unsupervised learning among decentralized data centers without requiring direct access to the distributed subgraphs. Specifically, FedGAD masks and reconstructs the neighborhood features to enhance the knowledge of node representations. Considering the data diversity across distributed clients, we also design a cross-clients' node representation module that enables nodes to reconstruct neighbors by leveraging information from other clients. Furthermore, we use a multiscale contrastive learning function, which includes both structure-level and contextual-level learning functions, to detect graph anomalies in the condition that subgraphs located at different clients show imbalanced data distributions. Experimental results on seven benchmark datasets demonstrate the superior performance of FedGAD compared with baseline methods, verifying its capability of improving GAD performance.

Knowledge distillation and teacher-student learning in medical imaging: Comprehensive overview, pivotal role, and future directions.

Foundations of Artificial Intelligence in Hepatology: What a Clinician Needs to Know.

The international legal framework for the protection of data privacy in health care

A verifiable and efficient chained federated learning scheme for privacy protection

Enhancing usability in face privacy protection via vision-language guided diffusion model

Conventional cross-subject electroencephalogram (EEG) signal identification frameworks require centralized aggregation of all subjects' data for feature extraction, which inherently poses substantial risks of data privacy breaches. In response to this critical issue, the present study delves into the classification of steady-state visual evoked potential (SSVEP) signals with an emphasis on data privacy preservation. First, we design a federated learning framework (FedGF) consisting of a central server and multiple clients, where the server generates global features and coordinates distributed training across clients, while retaining subject-specific raw data locally to ensure privacy protection. Then, to enhance model generalizability, FedGF employs data-free knowledge distillation (DFKD) to achieve knowledge transfer across clients through global feature learning. Extensive experiments on two public datasets (Dataset 1 'session01' and 2 'session02') and one private dataset (Dataset 3) demonstrate the superiority of the proposed method over baseline approaches, achieving performance improvements of 0.52%, 0.65%, and 0.53%, respectively. Finally, we develop a novel smart soft gripper with thermochromic capabilities and seamlessly integrate it with the trained network, demonstrating robust performance in daily grasping tasks. The source code is available at https://github.com/raow923/FedGF.

Information privacy protection in malicious node detection in wireless sensor networks

Privacy Protection Algorithms in Mobile Network Design and Innovation: Enhancing Security in Next-Generation Wireless Communication

With the popularity of personal devices, there are abundant valuable face image datasets in the industry, which provides opportunities for the development of visual models. However, privacy concerns related to identity sensitive information hinder face datasets sharing. Despite existing works dedicated to removing identity sensitive information from images, they either lack provable privacy guarantees or compromise crucial face dataset utilities, e.g., identity correlation and image naturalness. To overcome these weaknesses, we propose a novel face dataset publication scheme that protects face images by obfuscating face features. The obfuscated features still retain a certain level of correlation, allowing the protected dataset to be used for training. In the process of obfuscating the features, we design a novel metric differential privacy mechanism, which can enhance the correlation between features while ensuring privacy. Furthermore, we construct a latent diffusion model with identity and attribute as inputs to improve the naturalness of generated images. Extensive experimental results and theoretical analysis demonstrate our scheme significantly outperforms existing works in providing privacy protection while maintaining high dataset utility for downstream tasks.

Assistant-Based Integrity Auditing Scheme With Privacy Protection Function for Cloud Storage

Towards to real world vehicle privacy protection: A new dataset and benchmark

A labeled ophthalmic ultrasound dataset with medical report generation based on cross-modal deep learning.

Automated negotiation in multi-agent electronic commerce environments relies heavily on efficient and reliable matching mechanisms to connect negotiation participants. However, existing matching protocols often fail to ensure transaction security and user data privacy, while also lacking adaptability to dynamic negotiation contexts. To address these challenges, this study proposes a privacy-enhanced multi-agent matching optimization framework that integrates trust evaluation, privacy protection, and adaptive decision-making. First, a trust-based negotiation relationship network is constructed through complex network analysis to establish a secure and trustworthy negotiation environment. Second, a privacy-enhanced automated negotiation protocol is developed, employing the cumulative distribution function to transform sensitive data into probabilistic representations, thereby safeguarding user privacy without compromising data availability. Finally, a reinforcement learning algorithm is incorporated to optimize the matching process dynamically, using satisfaction as the reward function to achieve efficient and Pareto-optimal results. A series of experiments verify the framework’s effectiveness, demonstrating significant improvements in system robustness, adaptability, and matching accuracy. This study aims to provide a comprehensive solution that integrates trust network modeling, privacy protection, and adaptive matching optimization, serving as a valuable reference for the development of secure and intelligent automated negotiation platforms.

Banking and online financial service providers face significant challenges due to financial fraud. Traditional frauddetection methods are often inadequate because of imbalanced datasets, limited interpretability, and privacy concerns involving confidential customer information. This paper presents an explainable AI–based system for financial fraud detection designed to address these issues. The system employs the Light Gradient Boosting Machine (LightGBM) as the primary model, combined with SMOTE oversampling to mitigate class imbalance. Privacy is maintained by anonymizing sensitive features, including Personally Identifiable Information (PII), by temporarily adding and later removing attributes such as name_email_similarity before model training. Model transparency is achieved through SHAP (Shapley Additive Explanations), which offers featurelevel interpretability for fraud predictions. The system is implemented as a web-based interactive dashboard using the Flask framework, enabling users to upload datasets, perform fraud detection, adjust detection sensitivity (via threshold tuning), and download a detailed fraud report. When evaluated on a real-world dataset, the system achieved an overall accuracy of 98.5%, an ROC-AUC of 0.89, improved privacy preservation, and enhanced interpretability through SHAP. The proposed solution provides a practical end-to-end framework that balances accuracy, transparency, and privacy protection, making it suitable for banking and fintech fraud-detection applications.

This article presents new EU Regulation 2024/1689/EU (the ‘AI Act’), which establishes general rules for the development, marketing, and use of AI systems within the EU. Therefore, from 2 February 2025 onwards, the use and marketing of AI systems deemed particularly dangerous under the criteria set out in Chapter II of the new AI Act will be banned across the EU. Failure to comply may result in substantial administrative fines, with a maximum penalty of either €35 million or 7% of a company’s total worldwide turnover for the previous year. These relatively new experiences with AI-based products prompt lawyers to re-examine well-established legal concepts and to face challenges associated with the use of AI systems for law enforcement purposes. While AI systems employed in law enforcement can be beneficial, they may also interfere with people’s fundamental rights, most notably the rights to non-discrimination, freedom of expression, human dignity, and the protection of personal data and privacy. Therefore, the author seeks to identify the challenges that still lie ahead for legislators and courts in applying the AI Act. In analysing these issues, the author employs research methods relevant to legal studies, including formal-legal and legal-comparative approaches.

This study examines the formation of filter bubbles in social media ecosystems and analyses how they contribute to the intensification of political division. Drawing on a range of academic sources, the research examines the consequences of filter bubbles on key events such as the Arab Spring and Brexit, highlighting their impact on political mobilisation and communication. It also explores the complex relationship between social media, politics and law, and examines issues related to government control. Suggested measures to reduce polarisation include increasing the diversity of social networks and implementing policy reforms based on empirical evidence. However, balancing political sway, the right to self-expression and privacy protection remains a challenging task. This article emphasises the importance of multidisciplinary work in understanding and addressing the complex dynamics of the impact of social media on politics and society. It also calls for international cooperation in establishing legal structures to regulate the digital public sphere, guaranteeing democratic responsibility and social cohesion in the age of digitalisation.

The increasing size of data has created a pressing need for protection of communication and data privacy, spurring significant interest in quantization. This article proposes a novel scheme for variance reduced correlated quantization that is designed for data with bounded support and distributed mean estimation. Our method achieves a theoretical reduction in the mean square error for fixed and randomized designs compared to the correlated quantization method under different levels and dimensions scenarios. Several synthetic data experiments were conducted to illustrate the effectiveness of the approach and to provide a reliable approximation of the reduced mean square error based on the theory. The proposed method was also applied to real-world data in different learning tasks, which yielded promising results. Supplementary materials for this article are available online.

Security and Confidentiality of patient information are important in the modern healthcare system. Patient information is often stored on digital platforms through digital health records, telemedicine, and remote monitoring. The proposed work presents a cryptographic authentication framework for healthcare monitoring that uses AES-256 and Virtual Password Authentication(VPF) to protect sensitive data. The Virtual Password Function (VPF) is a little trick that combines a secret function with a code booking technique. This technique prevents unauthorized users from compromising security. It mitigates password-based attacks. Patient data is stored in a completely encrypted way to meet healthcare privacy mandates. The proposed system was developed in Java for encryption and matching authentication of processes. The implementation uses AES-256 encryption to safeguard patient data. It includes custom authentication logic for managing virtual passwords. The cloud uses encrypted end-to-end patient information and stores it in MySQL. The scalable and maintainable front-end web interface and backend control logic are developed using Java JSP Servlet. The framework provides secure, adequate protection of sensitive healthcare data in digital health ecosystems by leveraging strong encryption and adaptive authentication. As shown by experimental results and security analysis, the proposed model is effective for healthcare applications requiring high-level security. It offers relatively low execution, processing, key generation, and encryption/decryption times, alongside enhanced security.