Exposing Vulnerabilities in RL: A Novel Stealthy Backdoor Attack through Reward Poisoning
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866912734352769024 |
|---|---|
| author | Zhang, Bokang Lu, Chaojun Li, Jianhui Wu, Junfeng |
| author_facet | Zhang, Bokang Lu, Chaojun Li, Jianhui Wu, Junfeng |
| contents | Reinforcement learning (RL) has achieved remarkable success across diverse domains, enabling autonomous systems to learn and adapt to dynamic environments by optimizing a reward function. However, this reliance on reward signals creates a significant security vulnerability. In this paper, we study a stealthy backdoor attack that manipulates an agent's policy by poisoning its reward signals. The effectiveness of this attack highlights a critical threat to the integrity of deployed RL systems and calls for urgent defenses against training-time manipulation. We evaluate the attack across classic control and MuJoCo environments. The backdoored agent remains highly stealthy in Hopper and Walker2D, with minimal performance drops of only 2.18 % and 4.59 % under non-triggered scenarios, while achieving strong attack efficacy with up to 82.31% and 71.27% declines under trigger conditions. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_22415 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Exposing Vulnerabilities in RL: A Novel Stealthy Backdoor Attack through Reward Poisoning Zhang, Bokang Lu, Chaojun Li, Jianhui Wu, Junfeng Cryptography and Security Reinforcement learning (RL) has achieved remarkable success across diverse domains, enabling autonomous systems to learn and adapt to dynamic environments by optimizing a reward function. However, this reliance on reward signals creates a significant security vulnerability. In this paper, we study a stealthy backdoor attack that manipulates an agent's policy by poisoning its reward signals. The effectiveness of this attack highlights a critical threat to the integrity of deployed RL systems and calls for urgent defenses against training-time manipulation. We evaluate the attack across classic control and MuJoCo environments. The backdoored agent remains highly stealthy in Hopper and Walker2D, with minimal performance drops of only 2.18 % and 4.59 % under non-triggered scenarios, while achieving strong attack efficacy with up to 82.31% and 71.27% declines under trigger conditions. |
| title | Exposing Vulnerabilities in RL: A Novel Stealthy Backdoor Attack through Reward Poisoning |
| topic | Cryptography and Security |
| url | https://arxiv.org/abs/2511.22415 |