All You Need is "Leet": Evading Hate-speech Detection AI
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866916751364587520 |
|---|---|
| author | Kahu, Sampanna Yashwant Ahuja, Naman |
| author_facet | Kahu, Sampanna Yashwant Ahuja, Naman |
| contents | Social media and online forums are increasingly becoming popular. Unfortunately, these platforms are being used for spreading hate speech. In this paper, we design black-box techniques to protect users from hate-speech on online platforms by generating perturbations that can fool state of the art deep learning based hate speech detection models thereby decreasing their efficiency. We also ensure a minimal change in the original meaning of hate-speech. Our best perturbation attack is successfully able to evade hate-speech detection for 86.8 % of hateful text. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_16263 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | All You Need is "Leet": Evading Hate-speech Detection AI Kahu, Sampanna Yashwant Ahuja, Naman Cryptography and Security Computation and Language Machine Learning K.6.5 Social media and online forums are increasingly becoming popular. Unfortunately, these platforms are being used for spreading hate speech. In this paper, we design black-box techniques to protect users from hate-speech on online platforms by generating perturbations that can fool state of the art deep learning based hate speech detection models thereby decreasing their efficiency. We also ensure a minimal change in the original meaning of hate-speech. Our best perturbation attack is successfully able to evade hate-speech detection for 86.8 % of hateful text. |
| title | All You Need is "Leet": Evading Hate-speech Detection AI |
| topic | Cryptography and Security Computation and Language Machine Learning K.6.5 |
| url | https://arxiv.org/abs/2505.16263 |