# Security and Privacy on Generative Data in AIGC: A Survey

**Authors:** Tao Wang, Yushu Zhang, Shuren Qi, Ruoyu Zhao, Zhihua Xia, Jian Weng  
**arXiv:** [2309.09435](https://arxiv.org/abs/2309.09435)  
**Date:** 2023-09-18  
**Summary page:** https://www.summarizepaper.com/en/paper/2309.09435v1/

## Key points

- Artificial intelligence-generated content (AIGC) has revolutionized the field of information technology.
- AIGC enables the generation of high-quality data that is difficult to distinguish from real data.
- The widespread use of generative data in cyberspace raises concerns about security and privacy.
- Academia and industry have developed countermeasures to enhance security and privacy in AIGC.
- The authors systematically review the security and privacy aspects of generative data in AIGC from an information security perspective.
- They analyze properties including privacy, controllability, authenticity, and compliance.
- Successful experiences with state-of-the-art countermeasures are highlighted.
- Valuable insights into safeguarding generative data are provided from an information security perspective.
- Open challenges and potential exploration directions for each property are summarized in the survey.

## Summary

The advent of artificial intelligence-generated content (AIGC) has revolutionized the field of information technology. AIGC enables the effortless generation of high-quality data that is difficult for the public to distinguish from real data. However, the widespread use of generative data in cyberspace has raised concerns about security and privacy, such as privacy leakages of individuals and media forgery for fraudulent purposes. To address these challenges, both academia and industry have focused on ensuring the trustworthiness of generative data by developing a series of countermeasures to enhance security and privacy in AIGC. In this survey, the authors systematically review the security and privacy aspects of generative data in AIGC from an information security perspective. They analyze various properties including privacy, controllability, authenticity, and compliance. Furthermore, they highlight successful experiences with state-of-the-art countermeasures that address each property effectively. By examining these countermeasures from an information security perspective, they provide valuable insights into safeguarding generative data. Finally, the authors summarize open challenges and potential exploration directions for each property discussed in this comprehensive survey. This survey offers a deeper understanding of the security and privacy issues associated with generative data in AIGC and serves as a valuable resource for researchers and practitioners working towards enhancing trustworthiness in artificial intelligence-generated content.