Private Seeds, Public LLMs: Realistic and Privacy-Preserving Synthetic Data Generation

Computer Science > Cryptography and Security [Submitted on 8 Apr 2026] Private Seeds, Public LLMs: Realistic and Privacy-Preserving Synthetic Data Generation Qian Ma, Sarah Rajtmajer Large language models (LLMs) have emerged as a powerful tool for synthetic data generation. A particularly important use case is producing synthetic replicas of private text, which requires carefully balancing privacy and utility. We propose Realistic and Privacy-Preserving Synthetic Data Generation (RPSG), which leverages privacy-preserving mechanisms, including formal differential privacy (DP); and private seeds, in particular text containing personal information, to generate realistic synthetic data. Comprehensive experiments against state-of-the-art private synthetic data generation methods demonstrate that RPSG achieves high fidelity to private data while providing strong privacy protection. Comments: 23 pages, 7 figures, 18 tables Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI) Cite as: arXiv:2604.07486 [cs.CR]   (or arXiv:2604.07486v1 [cs.CR] for this version)   https://doi.org/10.48550/arXiv.2604.07486 Focus to learn more Submission history From: Qian Ma [view email] [v1] Wed, 8 Apr 2026 18:26:34 UTC (9,374 KB) Access Paper: HTML (experimental) view license Current browse context: cs.CR < prev   |   next > new | recent | 2026-04 Change to browse by: cs cs.AI References & Citations NASA ADS Google Scholar Semantic Scholar Export BibTeX Citation Bookmark Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, Media Demos Related Papers About arXivLabs Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)