arXiv：cs.CL· Shuai Wu, Xue Li, Zhijun Wang, Bolun Liu, Weilin Cai, Zihao Su, Ran Wang·· 3 小时前AI 评分33
前沿图像生成模型与合成视觉证据的真实世界风险
Seeing Is No Longer Believing: Frontier Image Generation Models, Synthetic Visual Evidence, and Real-World Risk
AI 导读
一篇综述梳理截至2026年10月1日的公开模型文档、事件报告与治理资料，分析前沿图像生成模型如何让合成图像获得证据效力。研究区分厂商能力宣称、已记录事件、实验发现与潜在危害路径，并比较供应商限制、来源溯源系统、水印、平台标注与政策义务在功能上不同于对事件本身的独立验证。作者主张高风险决策需依赖经认证的源记录、可信渠道佐证与行动前的比例化审查。
正文
Abstract:Image generation systems can produce plausible photographs, readable documents, and consistent depictions of people and places. When these artifacts are presented as records of real events, they can influence decisions in news, finance, identity verification, medicine, and law. This narrative review examines selected public model documentation, incident reports, research, and governance sources available through 1 October 2026, with English and Chinese community material providing illustrative context. We distinguish vendor capability claims, documented incidents, experimental findings, and prospective harm pathways. The analysis connects realism, text rendering, reference consistency, editing, grounding, and production cost to the conditions under which synthetic images acquire evidentiary authority. Historical incidents illustrate these pathways; they do not establish misuse rates for current models. We compare provider restrictions, provenance systems, watermarking, platform labeling, and policy obligations, and explain how their functions differ from independent verification of a depicted event or transaction. The framework links artifact types and decision contexts to the functions of available controls. High-stakes decisions call for authenticated source records, corroboration through trusted channels, and proportionate review before action.
| Comments: | 24 pages, 13 figures. Revised review of image-generation capabilities, synthetic visual evidence, and governance |
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI) |
| ACM classes: | I.2.7; I.2.10; K.4.1; K.6.5 |
| Cite as: | arXiv:2604.24197 [cs.CL] |
| (or arXiv:2604.24197v3 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2604.24197 arXiv-issued DOI via DataCite |
Submission history
From: Shuai Wu [view email]
[v1]
Mon, 27 Apr 2026 08:59:40 UTC (584 KB)
[v2]
Mon, 6 Jul 2026 10:10:55 UTC (584 KB)
[v3]
Wed, 7 Oct 2026 10:47:56 UTC (163 KB)
来源：arXiv：cs.CL · arxiv.org