Credit card fraud poses a critical threat to global financial systems, with annual losses exceeding $485 billion, necessitating robust detection mechanisms that balance predictive accuracy with stringent data privacy requirements. Existing approaches face fundamental limitations: Centralised machine learning models compromise regulatory compliance due to sharing the training data and model with third parties, while standard federated learning (FL) remains vulnerable to gradient-based privacy attacks. This paper presents a novel privacy-preserving framework integrating Conditional Tabular Generative Adversarial Networks (CTGAN) with Differentially Private FL. Unlike existing federated GAN approaches that suffer from training instability, our framework maintains local CTGAN synthesis whilst federating only the downstream classifier, achieving superior synthetic data fidelity. The framework employs local synthetic data generation with formal differential privacy (DP) guarantees at both the generation and the federated training stages. Extensive evaluation on two benchmark fraud datasets partitioned across three simulated banking institutions demonstrates an F1-score of 0.9485 ± 0.0014 across 10 independent runs, retaining 95.2% of the centralised-real baseline whilst providing formal (ε, δ)-DP at privacy budget (ε = 3.0) and failure probability (δ = 10–5). Privacy evaluation against five state-of-the-art defences, confirms that membership-inference accuracy remains at chance level (0.48–0.51) across all tested configurations, and that model-inversion reconstruction error increases under DP. This work advances privacy-preserving financial analytics by providing a rigorously validated, regulatory-compliant framework enabling collaborative fraud detection across distributed banking networks without compromising institutional data sovereignty.

An average U.K. electricity bill is made up of at least 60% service charge, with approximately 22% related to network characteristics including distance charge. This makes distance and network constraints important factors in matching prosumers on any peer-to-peer energy trading platform as assessed in this article. To realize that, a platform—$VirtElect$, based on a double auction market is developed to support the matching interaction between prosumers. Case studies based on real microgrid data are used to verify the performance of the platform in demonstrating the potential of local energy consumption. The results show that it is possible to balance local energy generation and consumption, with little or no interaction with the utility grid. We also show that local energy trading is not only beneficial to the environment but also leads to a significant amount of cost savings of up to 45%, depending on the number of participants and their ratios on the platform.

Deep learning (DL) has been widely proposed for botnet attack detection in Internet of Things (IoT) networks. However, the traditional centralized DL (CDL) method cannot be used to detect the previously unknown (zero-day) botnet attack without breaching the data privacy rights of the users. In this article, we propose the federated DL (FDL) method for zero-day botnet attack detection to avoid data privacy leakage in IoT-edge devices. In this method, an optimal deep neural network (DNN) architecture is employed for network traffic classification. A model parameter server remotely coordinates the independent training of the DNN models in multiple IoT-edge devices, while the federated averaging (FedAvg) algorithm is used to aggregate local model updates. A global DNN model is produced after a number of communication rounds between the model parameter server and the IoT-edge devices. The zero-day botnet attack scenarios in IoT-edge devices is simulated with the Bot-IoT and N-BaIoT data sets. Experiment results show that the FDL model: 1) detects zero-day botnet attacks with high classification performance; 2) guarantees data privacy and security; 3) has low communication overhead; 4) requires low-memory space for the storage of training data; and 5) has low network latency. Therefore, the FDL method outperformed CDL, localized DL, and distributed DL methods in this application scenario.

Blockchain (BC) and artificial intelligence (AI) are often utilized separately in energy trading systems (ETSs). However, these technologies can complement each other and reinforce their capabilities when integrated. This article provides a comprehensive review of consensus algorithms (CAs) of BC and deep reinforcement learning (DRL) in ETS. While the distributed consensus underpins the immutability of transaction records of prosumers, the deluge of data generated paves the way to use AI algorithms for forecasting and address other data analytic-related issues. Hence, the motivation to combine BC with AI to realize secure and intelligent ETS. This study explores the principles, potentials, models, active research efforts and unresolved challenges in the CA and DRL. The review shows that despite the current interest in each of these technologies, little effort has been made at jointly exploiting them in ETS due to some open issues. Therefore, new insights are actively required to harness the full potentials of CA and DRL in ETS. We propose a framework and offer some perspectives on effective BC-AI integration in ETS.