On the Trustworthiness Landscape of State-of-the-art Generative Models: A Survey and Outlook

• Published in: International journal of computer vision Vol. 133; no. 7; pp. 4317 - 4348
• Main Authors: Fan, Mingyuan, Wang, Chengyu, Chen, Cen, Liu, Yang, Huang, Jun
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.07.2025; Springer Nature B.V
• Subjects: Accountability; Artificial Intelligence; Computer Imaging; Computer Science; Computer vision; Diffusion models; Generative artificial intelligence; Image Processing and Computer Vision; Investigations; Large language models; Pattern Recognition; Pattern Recognition and Graphics; Privacy; Special Issue on Large-Scale Generative Models for Content Creation and Manipulation; Trustworthiness; Vision; Privacy; Trustworthiness; Large language models; Diffusion models; Security; Responsibility; Fairness
• ISSN: 0920-5691; 1573-1405
• DOI: 10.1007/s11263-025-02375-w

Diffusion models and large language models have emerged as leading-edge generative models, revolutionizing various aspects of human life. However, their practical implementation has also exposed inherent risks, bringing to light their potential downsides and sparking concerns about their trustworthiness. Despite the wealth of literature on this subject, a comprehensive survey that specifically delves into the intersection of large-scale generative models and their trustworthiness remains largely absent. To bridge this gap, this paper investigates both long-standing and emerging threats associated with these models across four fundamental dimensions: 1) privacy, 2) security, 3) fairness, and 4) responsibility. Based on our investigation results, we develop an extensive survey that outlines the trustworthiness of large generative models. Following that, we provide practical recommendations and identify promising research directions for generative AI, ultimately promoting the trustworthiness of these models and benefiting society as a whole.