Multilingual Hidden Prompt Injection Attacks on LLM-Based Academic Reviewing
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866909977856180224 |
|---|---|
| author | Theocharopoulos, Panagiotis Kulkarni, Ajinkya -Doss, Mathew Magimai. |
| author_facet | Theocharopoulos, Panagiotis Kulkarni, Ajinkya -Doss, Mathew Magimai. |
| contents | Large language models (LLMs) are increasingly considered for use in high-impact workflows, including academic peer review. However, LLMs are vulnerable to document-level hidden prompt injection attacks. In this work, we construct a dataset of approximately 500 real academic papers accepted to ICML and evaluate the effect of embedding hidden adversarial prompts within these documents. Each paper is injected with semantically equivalent instructions in four different languages and reviewed using an LLM. We find that prompt injection induces substantial changes in review scores and accept/reject decisions for English, Japanese, and Chinese injections, while Arabic injections produce little to no effect. These results highlight the susceptibility of LLM-based reviewing systems to document-level prompt injection and reveal notable differences in vulnerability across languages. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_23684 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Multilingual Hidden Prompt Injection Attacks on LLM-Based Academic Reviewing Theocharopoulos, Panagiotis Kulkarni, Ajinkya -Doss, Mathew Magimai. Computation and Language Artificial Intelligence Large language models (LLMs) are increasingly considered for use in high-impact workflows, including academic peer review. However, LLMs are vulnerable to document-level hidden prompt injection attacks. In this work, we construct a dataset of approximately 500 real academic papers accepted to ICML and evaluate the effect of embedding hidden adversarial prompts within these documents. Each paper is injected with semantically equivalent instructions in four different languages and reviewed using an LLM. We find that prompt injection induces substantial changes in review scores and accept/reject decisions for English, Japanese, and Chinese injections, while Arabic injections produce little to no effect. These results highlight the susceptibility of LLM-based reviewing systems to document-level prompt injection and reveal notable differences in vulnerability across languages. |
| title | Multilingual Hidden Prompt Injection Attacks on LLM-Based Academic Reviewing |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2512.23684 |