# Safety and adversarial attacks

Area: [Vision-language-action models](https://mlsanity.com/vla/index.md). HTML: https://mlsanity.com/vla/safety-and-adversarial-attacks/

## Key facts

- Safety and adversarial attacks: 58 papers on arXiv since 2024 (data through 7 Oct 2026).
- 29 of its 58 mapped papers (50%) cite “Exploring the Adversarial Vulnerabilities of Vision-Language-Action Models in Robotics” (2024, arXiv 2411.13587).
- 18 of its 58 mapped papers (31%) cite “BadVLA: Towards Backdoor Attacks on Vision-Language-Action Models via Objective-Decoupled Optimization” (2025, arXiv 2505.16640).
- 9 new papers 17 Sep–7 Oct.

## Most cited here

- [Exploring the Adversarial Vulnerabilities of Vision-Language-Action Models in Robotics](https://mlsanity.com/paper/2411.13587/) (2024), cited by 29 papers here
- [BadVLA: Towards Backdoor Attacks on Vision-Language-Action Models via Objective-Decoupled Optimization](https://mlsanity.com/paper/2505.16640/) (2025), cited by 18 papers here
- [SafeVLA: Towards Safety Alignment of Vision-Language-Action Model via Constrained Learning](https://mlsanity.com/paper/2503.03480/) (2025), cited by 16 papers here
- [Adversarial Attacks on Robotic Vision Language Action Models](https://mlsanity.com/paper/2506.03350/) (2025), cited by 11 papers here
- [VLSA: Vision-Language-Action Models with Plug-and-Play Safety Constraint Layer](https://mlsanity.com/paper/2512.11891/) (2025), cited by 10 papers here
- [Manipulation Facing Threats: Evaluating Physical Vulnerabilities in End-to-End Vision Language Action Models](https://mlsanity.com/paper/2409.13174/) (2024), cited by 9 papers here
- [FreezeVLA: Action-Freezing Attacks against Vision-Language-Action Models](https://mlsanity.com/paper/2509.19870/) (2025), cited by 8 papers here
- [When Alignment Fails: Multimodal Adversarial Attacks on Vision-Language-Action Models](https://mlsanity.com/paper/2511.16203/) (2025), cited by 7 papers here
- [Model-agnostic Adversarial Attack and Defense for Vision-Language-Action Models](https://mlsanity.com/paper/2510.13237/) (2025), cited by 6 papers here
- [Goal-oriented Backdoor Attack against Vision-Language-Action Models via Physical Objects](https://mlsanity.com/paper/2510.09269/) (2025), cited by 5 papers here

## New (17 Sep–7 Oct)

- [TMT: Runtime Backdoor Detection for Vision-Language-Action Policies on Unseen Tasks](https://mlsanity.com/paper/2610.09462/) (2026-10-07)
- [Detect and Suppress: A Mechanistic Defense against Adversarial Patches in VLA Models](https://mlsanity.com/paper/2610.03498/) (2026-10-02)
- [WBAG: A Whole-Body and Attached-Geometry Safety Framework for Vision-Language-Action Manipulation](https://mlsanity.com/paper/2610.01083/) (2026-10-01)
- [Multi-Link Safety Filtering for VLA Policies Around Moving Hazards](https://mlsanity.com/paper/2609.40007/) (2026-09-30)
- [Exploiting Vulnerabilities: Universal Adversarial Attacks on Vision-Language-Action Models in Robotics](https://mlsanity.com/paper/2609.39178/) (2026-09-30)
- [Are Vision-Language-Action Models Robust to One-Step Observation Perturbations?](https://mlsanity.com/paper/2609.32550/) (2026-09-26)
- [SafeLoop: Risk-Aware Rollback for Vision-Language-Action Manipulation](https://mlsanity.com/paper/2609.26313/) (2026-09-22)
- [VLPSA: Vision-Language-Poisson-Safe Actions for Full-Body Safety of Learned Policies](https://mlsanity.com/paper/2609.22462/) (2026-09-18)
- [Beyond Patch Removal: Persistent Adversarial Effects in Vision-Language-Action Policies](https://mlsanity.com/paper/2609.19669/) (2026-09-17)