With the increasing use of artificial intelligence (AI) models, more attention is being paid to issues of trust and security in AI systems against various types of threats (evasion attacks, poisoning, membership inference, etc.). In this work, we focus on the task of graph node classification, highlighting it as one of the most complex. To the best of our knowledge, this is the first study exploring the relationship between defense methods for AI models against different types of threats on graph data. Our experiments are conducted on citation and purchase graph datasets. We demonstrate that, in general, it is not advisable to simply combine defense methods for different types of threats, as this can lead to severe negative consequences, including a complete loss of model effectiveness. Furthermore, we provide theoretical proof of the contradiction between defense methods against poisoning attacks on graphs and adversarial training.
Image Quality Assessment (IQA) models are increasingly relied upon to evaluate image quality in real-world systems -- from compression and enhancement to generation and streaming. Yet their adoption brings a fundamental risk: these models are inherently unstable. Adversarial manipulations can easily fool them, inflating scores and undermining trust. Traditionally, such vulnerabilities are addressed through data-driven defenses -- adversarial retraining, regularization, or input purification. But what if this is the wrong lens? What if robustness in perceptual models is not something to learn but something to design? In this work, we propose a provocative idea: robustness as an architectural prior. Rather than training models to resist perturbations, we reshape their internal structure to suppress sensitivity from the ground up. We achieve this by enforcing orthogonal information flow, constraining the network to norm-preserving operations -- and further stabilizing the system through pruning and fine-tuning. The result is a robust IQA architecture that withstands adversarial attacks without requiring adversarial training or significant changes to the original model. This approach suggests a shift in perspective: from optimizing robustness through data to engineering it through design.
1 December 2025 in Programming and Computer Software
With the increasing use of artificial intelligence (AI) models, more attention is being paid to trust and security of AI systems against threats of various types (evasion attacks, poisoning, membership inference, etc.). In this work, we focus on graph node classification, highlighting it as one of the most complex tasks. To the best of our knowledge, this is the first study that explores the relationship between defense methods for AI models against different threats on graph data. Our experiments are conducted on citation and purchase graph datasets. We demonstrate that, in general, it is not advisable to simply combine defense methods against different types of threats, as this can lead to severe negative consequences, including a complete loss of model effectiveness. Furthermore, we provide a theoretical proof of the contradiction between defense methods against poisoning attacks on graphs and adversarial training.