| title | SynCred-Bench: Benchmarking Synthetic Credibility in AI-Generated Visual Misinformation |
|---|---|
| authors | Junxiao Yang, Minghao Zhang, Xiaoce Wang, Haoran Liu, Shiyao Cui, Hongning Wang, Minlie Huang |
| date | 2026-06-02 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2606.03348 |
| upvotes | 1 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf |
| abstract | Recent generative models can now produce visual artifacts with realistic embedded text and layouts, creating a new misinformation threat: synthetic credibility. We introduce SYNCRED-Bench, a benchmark of 600 AI-generated misinformation images balanced across six credible-form categories and seven fine-grained circulation styles, together with FP450, a real-image negative set for measuring false positives. Extensive evaluation shows that existing systems remain unreliable: under a 5% false-positive-rate constraint, 15 MLLMs achieve only 10.5% true positive rate (TPR), open-source AIGC detectors achieve less than 5%, and commercial APIs reach 57.6%. Human annotators also struggled to identify synthetic credibility, reaching only 63% TPR. These findings establish synthetic credibility as a severe and underexplored visual misinformation challenge, and provide a benchmark for developing detectors that reason beyond superficial credibility cues. |
cite
@misc{yang2026syncred,
title={SynCred-Bench: Benchmarking Synthetic Credibility in AI-Generated Visual Misinformation},
author={Junxiao Yang and Minghao Zhang and Xiaoce Wang and Haoran Liu and Shiyao Cui and Hongning Wang and Minlie Huang},
year={2026},
eprint={2606.03348},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2606.03348}
}