CVPR Findings2026
Adversarial Agents: Black-Box Evasion Attacks with Reinforcement Learning
Kyle Domico, Jean-Charles Noirot Ferrand, Ryan Sheatsley, Eric Pauley, Josiah Hanna, Patrick McDaniel
Abstract
Attacks on machine learning models have been extensively studied through stateless optimization. In this paper, we demonstrate how a reinforcement learning (RL) agent can learn a new class of attack algorithms that generate adversarial samples. Unlike traditional adversarial machine learning (AML) methods that craft adversarial samples independently, our RL-based approach retains and exploits past attack experience to improve the effectiveness and efficiency of future attacks. We formulate adversarial sample generation as a Markov Decision Process and evaluate RL's ability to (a) learn effective and efficient attack strategies and (b) compete with state-of-the-art AML. On two image classification benchmarks, our agent increases attack success rate by up to 13.2% and decreases the average number of victim model queries per attack by up to 16.9% from the start to the end of training. In a head-to-head comparison with state-of-the-art image attacks, our approach enables an adversary to generate adversarial samples with 17% more success on unseen inputs post-training. From a security perspective, this work demonstrates a powerful new attack vector that uses RL to train agents that attack ML models efficiently and at scale.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
Lineage
- Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attacks cites Towards Evaluating the Robustness of Neural Networks
- HopSkipJumpAttack: A Query-Efficient Decision-Based Attack cites Towards Evaluating the Robustness of Neural Networks
Builds on4
- An Image is Worth 16x16 Words: Transformers for Image Recognition at ScaleAlexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn et al.ICLR 2021 · 21,477 citations
- Towards Evaluating the Robustness of Neural NetworksNicholas Carlini, David A. WagnerS&P 2017 · 9,786 citations
- Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attacksFrancesco Croce, Matthias HeinICML 2020 · 2,337 citations
- HopSkipJumpAttack: A Query-Efficient Decision-Based AttackJianbo Chen, Michael I. Jordan, Martin J. WainwrightS&P 2020 · 797 citations
Related papers
- Adversarial Policy Training against Deep Reinforcement LearningXian Wu, Wenbo Guo, Hua Wei, Xinyu XingUSENIX Security 2021 · 19 citations
- Experience Speaks Louder: Black-box Hard-label Adversarial Attack through Reinforcement LearningYilun Jin, Kun Zhu, Feng Tang, Jiaxuan Shi et al.KDD 2025 · 1 citation
- Who Is the Strongest Enemy? Towards Optimal and Efficient Evasion Attacks in Deep RLYanchao Sun, Ruijie Zheng, Yongyuan Liang, Furong HuangICLR 2022 · 82 citations
- Reinforcement Learning-based Adversarial Attacks on Object Detectors using Reward ShapingZhenbo Shi, Wei Yang, Zhenbo Xu, Zhidong Yu et al.ACM MM 2023 · 3 citations
- Stealthy and Efficient Adversarial Attacks against Deep Reinforcement LearningJianwen Sun, Tianwei Zhang, Xiaofei Xie, Lei Ma et al.AAAI 2020 · 141 citations
