Could AI Trace and Explain the Origins of AI-Generated Images and Text?
Hongchao Fang, Yixin Liu, Jiangshu Du, Can Qin, Ran Xu, Feng Liu, Lichao Sun, Dongwon Lee, Lifu Huang, Wenpeng Yin · 2025
Read Could AI Trace and Explain the Origins of AI-Generated Images and Text? by Hongchao Fang, Yixin Liu, Jiangshu Du, Can Qin, Ran Xu, Feng Liu, Lichao Sun, Dongwon Lee, Lifu Huang, Wenpeng Yin (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-04279.txt.
Description
Hongchao Fang and co-authors ask, in a 2025 paper, whether AI systems can trace fake images and text back to the models that made them, and explain the verdict. Machine-made content now floods public life. Abusers forge images that break rules, and lazy reviewers pass off model-written reviews as their own.
Earlier work mostly detected fakes, sometimes naming a source model. A systematic, fine-grained comparison was missing. The study spans images versus text, fully versus partly generated pictures, and benign versus malicious uses. It also tests whether a leading multimodal model can justify its attributions. Provenance tools that explain themselves would serve platforms, courts, and journals better than bare detectors.
Catalog record
| CREATOR | Hongchao Fang, Yixin Liu, Jiangshu Du, Can Qin, Ran Xu, Feng Liu, Lichao Sun, Dongwon Lee, Lifu Huang, Wenpeng Yin |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 40 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-04279