StepGuard introduces a step-level guard model that audits and checks tool actions before execution, addressing security risks in LLM-based agents. It is trained using StepGen, an automatic engine that generates safe and unsafe trajectories, and employs Balance-GRPO to dynamically balance learning between safe and unsafe actions. Experiments show StepGuard achieves high accuracy comparable to GPT-5.4 and significantly reduces attack success rates while minimally impacting utility.
By Zhijie Zheng, Yu Li, Chen Qian, Yuqian Fu, Yanwei Fu, Lu Sheng, Jing Shao, Dongrui Liu
arXiv:2606. 18223v1 Announce Type: cross Abstract: With sophisticated cyber-attacks becoming increasingly prevalent, modern networks require intelligent autonomous cyber-defense agents trained via Reinforcement Learning (RL).
By Ankita Samaddar, Sandeep Neema, Daniel Balasubramanian, Xenofon Koutsoukos
arXiv:2606. 29867v1 Announce Type: cross Abstract: Deep Reinforcement Learning (DRL) has achieved significant success in robotics and autonomous systems, yet remains vulnerable to adversarial perturbations that can severely degrade performance.
By Adithya Mohan, Daniel Kriegl, Torsten Sch\"on
arXiv:2603. 13026v2 Announce Type: replace Abstract: Prompt injection poses serious security risks to real-world LLM applications, particularly autonomous agents.
By Chenlong Yin, Runpeng Geng, Yanting Wang, Jinyuan Jia
arXiv:2505. 19532v2 Announce Type: replace Abstract: The current state-of-the-art backdoor attacks against Reinforcement Learning (RL) rely upon unrealistically permissive access models, that assume the attacker can read (or even write) the victim's policy parameters, observations, or rewards.
By Shijie Liu, Andrew C. Cullen, Paul Montague, Sarah Erfani, Benjamin I. P. Rubinstein
arXiv:2607. 06643v1 Announce Type: cross Abstract: Backdoor attacks severely threaten large-scale AI models.
By Issam Seddik, Sami Souihi, Mohamed Tamaazousti, Sara Tucci Piergiovanni
arXiv:2506. 22423v2 Announce Type: replace Abstract: Unmanned Aerial Vehicles (UAVs) depend on onboard sensors for perception, navigation, and control.
By Pritam Dash, Ethan Chan, Nathan P. Lawrence, Karthik Pattabiraman
arXiv:2608. 19836v1 Announce Type: cross Abstract: Probabilistic shielding is a technique for safe reinforcement learning (RL).
By Astrid Horn Brorholt (Aalborg University, Aalborg, Denmark), Maris F. L. Galesloot (Radboud University, Nijmegen, Netherlands), Nils Jansen (Radboud University, Nijmegen, Netherlands), Kim Guldstrand Larsen (Aalborg University, Aalborg, Denmark), Christian Schilling (Aalborg University, Aalborg, Denmark)
arXiv:2608. 04317v1 Announce Type: cross Abstract: Autonomous cyber defense systems based on Deep Reinforcement Learning (DRL) have attracted significant research attention, yet remain evaluated almost exclusively against static, heuristic red agents, leaving their robustness against adaptive threats critically understudied.
By Ryozo Masukawa, Ian Bryant, Armita Kazeminajafabadi, Sanggeon Yun, Hyunwoo Oh, SungHeon Jeong, Nathaniel D. Bastian, Mahdi Imani, Mohsen Imani
arXiv:2503. 01734v3 Announce Type: replace-cross Abstract: Attacks on machine learning models have been extensively studied through stateless optimization.
By Kyle Domico, Jean-Charles Noirot Ferrand, Ryan Sheatsley, Eric Pauley, Josiah Hanna, Patrick McDaniel
Autonomous cyber defense systems based on Deep Reinforcement Learning (DRL) have attracted significant research attention, yet remain evaluated almost exclusively against static, heuristic red agents, leaving their robustness against adaptive threats critically understudied. Meanwhile, recent advances in Reinforcement Learning with Verifiable Rewards (RLVR) have improved LLM reasoning, but their integration into cybersecurity remains elusive due to the absence of suitable benchmark environments and interaction datasets.
The paper presents a robust multi‑agent reinforcement learning framework for small unmanned aircraft systems (sUAS) to maintain separation assurance when GPS data is degraded or spoofed. By modeling state observation corruption as a zero‑sum game, the authors derive a closed‑form adversarial perturbation that eliminates iterative inner optimization and can be evaluated in linear time. Integrating this perturbation into a policy‑gradient MARL algorithm yields a counter‑policy that achieves near‑zero collision rates in high‑density simulations even with up to 35% observation corruption, outperforming non‑adversarial baselines.
By Alex Zongo, Filippos Fotiadis, Ufuk Topcu, Peng Wei