The field of differential privacy and synthetic data generation is moving towards more realistic and practical applications. Researchers are developing new methods to extend differential privacy guarantees to more realistic adversaries and to improve the utility of synthetic data. One of the key directions is to provide more nuanced and high-probability bounds for differential privacy mechanisms, which can help to better understand the privacy leakage in various attack settings. Another direction is to develop more efficient and scalable algorithms for synthetic data generation, which can maintain data sovereignty and comply with regulatory standards. Noteworthy papers in this area include 'Beyond the Worst Case: Extending Differential Privacy Guarantees to Realistic Adversaries', which presents a flexible framework for computing high-probability guarantees for DP mechanisms, and 'SynthGuard: Redefining Synthetic Data Generation with a Scalable and Privacy-Preserving Workflow Framework', which introduces a framework for ensuring computational governance in synthetic data generation. Additionally, 'Crypto-Assisted Graph Degree Sequence Release under Local Differential Privacy' presents an efficient framework for releasing degree sequences under local differential privacy, and 'Towards High Supervised Learning Utility Training Data Generation: Data Pruning and Column Reordering' proposes a novel pipeline for integrating data-centric techniques into tabular data synthesis.