| title | No Hidden Prompts Needed! You Can Game AI Peer Review with Presentation-Only Revisions |
|---|---|
| authors | Xu Yang, Zhizhou Sha, Junbo Li, Jian Yu, Yifan Sun, Matthew Zhao, Jinrui Fang, Xinyue Guo, Yining Wu, Xu Hu, Yifu Luo, Qiang Liu, Zhangyang Wang |
| date | 2026-06-11 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2606.13044 |
| affiliation | University of Texas at Austin |
| upvotes | 10 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf code |
| abstract | As AI-generated reviews move from experimental tools into peer-review infrastructure, most robustness concerns have focused on explicit attacks such as hidden instructions and prompt injection. We study a harder and more policy-relevant failure mode: no hidden text, no prompt injection, and no changes to methods, experiments, figures, equations, proofs, or numerical results. The attacker modifies only presentation-level content, such as the abstract, contribution framing, related work, discussion, and narrative structure. We introduce adversarial repackaging: a closed-loop attack that uses AI-reviewer feedback to search for presentation-level revisions while keeping the scientific evidence fixed. Across three mainstream AI reviewers, adversarial repackaging achieves a 75.1% attack success rate and a mean score gain of +1.21/10. The effect is not explained by ordinary prose polishing. We also reveal that strategies that change how the reviewer interprets the paper, such as related-work repositioning and analytical discussion expansion, substantially outperform surface edits such as local polishing, table formatting, and algorithm boxes. Our analysis reveals two deeper structural failure modes. First, AI reviewers are easier to impress than to convince: highlighting strengths reliably increases perceived merit, while attempts to dissolve weaknesses frequently backfire. Second, AI reviewers can confuse the appearance of addressing a limitation with actually resolving it, allowing unchanged evidence to be reinterpreted as stronger scientific contribution. These results show that the deployment risk is not only malicious hidden instructions, but the emergence of paper presentation itself as an optimization surface. We release a contamination-free rolling benchmark and attack framework for testing whether AI reviewers remain anchored to scientific content under presentation-only edits. |
cite
@misc{yang2026no,
title={No Hidden Prompts Needed! You Can Game AI Peer Review with Presentation-Only Revisions},
author={Xu Yang and Zhizhou Sha and Junbo Li and Jian Yu and Yifan Sun and Matthew Zhao and Jinrui Fang and Xinyue Guo and Yining Wu and Xu Hu and Yifu Luo and Qiang Liu and Zhangyang Wang},
year={2026},
eprint={2606.13044},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2606.13044}
}