The field of graph neural networks is moving towards addressing the challenges of anomaly detection and fairness. Researchers are exploring innovative approaches to improve the performance of graph anomaly detection models, including the use of spectral neural networks, causal edge separation, and contrastive learning. Additionally, there is a growing focus on ensuring fairness in graph neural networks, with methods being developed to mitigate bias and promote unbiased communities. Notable papers in this area include: From Pixels to Graphs, which evaluates the effectiveness of various image-to-graph transformation approaches for graph-level anomaly detection. CRoC, which proposes a framework for training GNNs with limited labeled data and achieves state-of-the-art performance on several GAD datasets. Improving Fairness in Graph Neural Networks via Counterfactual Debiasing, which presents a novel approach for mitigating bias in GNNs using counterfactual data augmentation. Addressing Graph Anomaly Detection via Causal Edge Separation and Spectrum, which analyzes the spectral distribution of nodes and proposes a spectral neural network for anomaly detection. Enhancing Contrastive Link Prediction With Edge Balancing Augmentation, which provides a theoretical analysis for contrastive learning on link prediction and proposes a new graph augmentation approach. Let's Grow an Unbiased Community, which proposes a framework for guiding the fairness of graphs via new links. GRASPED, which proposes a graph autoencoder with spectral encoder and decoder for node anomaly detection.