Adversarial Attacks

Follow this topic Ask about this topic

Adversarial attacks study how deliberately chosen inputs, training data, model parameters, or system interactions can make AI systems fail or behave as intended by an attacker. Common approaches include gradient-based perturbations, black-box query and evolutionary search, generative or semantic changes, physical patches and sensor interference, and poisoning or backdoors. Work spans image, language, audio, graph, and time-series models as well as federated learning, recommenders, robotics, and AI agents, with evaluations on benchmarks and increasingly realistic deployment settings. A central tension is that attacks must be effective and stealthy under practical constraints, while defenses and benchmarks must keep pace with adaptive threats and distinguish digital robustness from real-world resilience.

Latest papers

451 papers54 papers in the last four weeks, up 69% on the four weeks before. 0.3% of all new papers.

PASA-KV: Attacking Action-Facing Visual Routes in World Action Models
Oct 29, 2026Zhongbo Wu, Jiakai Huang, Zheng Zhang +6Adversarial Attacks on World Models Adversarial Attacks
Shanghai Jiao Tong University · Shanghai Innovation Institute · ACE Robotics +2
FlipJev: Fast Black-Box Red Teaming Finds Universal Adversarial Suffixes for Typed Decision Models
Oct 28, 2026Xiaogeng Liu, Xinyan Wang, Chaowei XiaoAdversarial Attacks Adversarial Text Attacks
Johns Hopkins University · University of Wisconsin–Madison
A Separation Principle for Adversarial Attacks and Defense of Graph Neural Networks
Oct 28, 2026Farid Mafi, Saber Jafarpour, Mohammad PiraniAdversarial Attacks Graph Neural Networks
University of Waterloo, Waterloo, ON, Canada · University of Colorado Boulder, Boulder, CO, USA · University of Ottawa, Ottawa, ON, Canada
When Is a Relation Worth Trusting? Manipulation Costs, Exposure, and the Trust Trap in Multi-Relational Graph Learning
Oct 27, 2026Vincess Dongmo, Astride Melvin Fokam NinyimGraph Robustness Adversarial Attacks
SYNTHI-AI, Cameroon · Universit´e du Québec à Chicoutimi (UQAC), Canada
Quantifying and Prioritizing Stealthy Source-Manipulation Threats Against Multi-Source Anomaly Detectors
Oct 27, 2026Indraveni Chebolu, Seema Panghal, Arnab Mallick +1Adversarial Attacks Anomaly Detection
Information Security Group, Centre for Development of Advanced Computing, Hyderabad, Telangana, India
The Target Agent Is Its Own Best Critic but a Poor Proposer: Self-Improved Red Teaming
Oct 26, 2026Mintong Kang, Stephen Rawls, Melissa Kazemi Rad +2LLM Red Teaming Automated Red Teaming
University of Illinois Urbana-Champaign · Capital One
Attack logics, not outputs: Concept-based Property Attacks for Shortcut Elimination in Neural Networks
Oct 26, 2026Raik Dankworth, Felix Mächtle, Thomas Eisenbarth +1Adversarial Attacks Adversarial Robustness
Institute of Software Engineering and Programming Languages University of Luebeck · Institute of IT Security University of Luebeck
Data-Efficient Backdoor Attacks on Code Models via Trigger-Neutralized Counterfactual Scoring
Oct 23, 2026Yu-Yu Lee, Ming-Feng Tsai, Tomohiro Morikawa +1Backdoor Attacks LLM Backdoor Attacks
Department of Computer Science and Engineering, National Sun Yat-sen University, Kaohsiung 804, Taiwan · Department of Computer Science and Engineering and the Security Research Center, National Sun Yat-sen University, Kaohsiung 804, Taiwan · University of Hyogo, Kobe, Japan +1
Towards Effective and Transferable Physical Camouflage against Multi-View BEV-based 3D Perception in Autonomous Driving
Oct 23, 2026Linye Lyu, Jiawei Zhou, Daojing He +1Adversarial Attacks Physical Adversarial Attacks
MIG-COW: Transferable Adversarial Attacks on Deepfake Detectors via Gradient Decomposition
Oct 22, 2026Wonjune Seo, Joonhyuk Baek, Yeseong Jung +1Adversarial Attacks Transfer-Based Adversarial Attacks
Ulsan National Institute of Science and Technology, Ulsan, Republic of Korea
Towards Adversarially Robust Multigraph Neural Networks for Financial Crime Analysis
Oct 22, 2026Dennis Heijmans, H. Çağrı Bilgi, Rui Wang +1Adversarial Attacks Adversarial Robustness
Delft University of Technology Delft, The Netherlands
On the Adversarial Vulnerability of Neural Receivers with Unrolled LDPC Decoders
Oct 21, 2026Aakash Saini, Jafar Aco Mohammadi, Wolfgang GerstackerAdversarial Attacks Neural Network Robustness
Nokia, Munich, Germany · Institute for Digital Communications, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany
How Much Do $\ell_1$ Attacks Need Concentrated Perturbations?
Oct 20, 2026Melika Shirian, Kianoosh Vadaei, Arshia HemmatAdversarial Attacks Adversarial Perturbations
University of Cambridge
Bridging Digital Deception and Physical Fidelity: A Spatio-Temporal Attack on Cognitive Radars
Oct 20, 2026Tongrun Xing, Yongjun Dai, Xuan Liao +4Adversarial Attacks Physical Adversarial Attacks
College of Computer Science and Technology, National University of Defense Technology, Changsha 410073, China. · College of Advanced Interdisciplinary Studies, National University of Defense Technology, Changsha 410073, China. · College of Science, National University of Defense Technology, Changsha 410073, China. +1
Detecting AI Training in Datacentre Power
Oct 19, 2026Tom Kimpson, Mauricio Baker, Emlyn GrahamAdversarial Attacks
School of Mathematics and Statistics, The University of Melbourne, Parkville, VIC 3010, Australia · Machine Alignment, Transparency & Security (MATS) · Department of Engineering Science, University of Oxford, Oxford, UK
Vulnerability Analysis for Safe Multi-Agent Reinforcement Learning
Oct 19, 2026Taozheng Zhu, Shixiong Jiang, Fanxin KongAdversarial Attacks Adversarial MARL
University of Notre Dame Notre Dame, Indiana, USA
Mind Your Boundaries: Boundary-Aware Model Obfuscation against Cross-node Attack
Oct 19, 2026Yuyang Zhang, Yihui Jin, Kangjie Chen +2Adversarial Robustness Adversarial Attacks
Tricking Synthetic Image Detectors with Amortized Generative Attacks
Oct 18, 2026Mohamed Awad, Khaled Elhady, Yasser Rohaim +2Adversarial Attacks on AI-Generated Image Detectors Adversarial Attacks
Mohamed bin Zayed University of Artificial Intelligence Abu Dhabi, United Arab Emirates
Robust Direct Preference Optimization under Preference Label-Flip Attacks
Oct 18, 2026Chenye Yang, Weiyu Xu, Lifeng LaiDirect Preference Optimization Adversarial Robustness
Department of Electrical and Computer Engineering, University of California, Davis, Davis, CA 95616, USA · Department of Electrical and Computer Engineering, University of Iowa, Iowa City, IA 52242, USA
Enhancing Visual Task-Redirection Attacks on Vision-Language-Action Models via Closed-Loop Optimization
Oct 16, 2026Chunjin Ye, Guanying Chen, Jianjie Huang +6Adversarial Attacks on VLA Models Adversarial Attacks
Sun Yat-sen University, Shenzhen 518107, China · Zhongguancun Academy, Beijing 100094, China · Wuhan University, Wuhan 430072, China +3
A Valid Watermark Is Not a Valid Identity: Spatial-Spectral Adversarial Face Evasion with Watermark Preservation
Oct 14, 2026Jefferson David Rodriguez Chivata, Davide Ghiani, Simone Maurizio La Cava +3Adversarial Attacks Black-Box Adversarial Attacks
University of Cagliari, Piazza d’Armi I, 09123 Cagliari, Italy
AIRRI: Towards a Human-Readable, Machine-Resistant Reading Interface
Oct 13, 2026Emory Salaun, Syed Ayon, Kaleel Mahmood +1Educational Assessment Optical Character Recognition
Department of Computer Science and Statistics University of Rhode Island Kingston, Rhode Island, USA
Mitigating Black-Box Score-Based Query Attacks via Gradient Confusion
Oct 12, 2026Ali Javeri, Javad Hezareh, Radin Shahdaei +1Adversarial Robustness Adversarial Attacks
Department of Computer Engineering, Sharif University of Technology, Tehran, Iran
SSM-Attack: Learning Where to Perturb for Black-Box Face Privacy
Oct 12, 2026Tunahan Parlayıcı, Erchan Aptoula, Yücel SaygınAdversarial Perturbations Adversarial Attacks
Computer Vision and Pattern Analysis Laboratory (VPALab), Faculty of Engineering and Natural Sciences, Sabanci University, Istanbul, Türkiye · Faculty of Engineering and Natural Sciences, Sabanci University, Istanbul, Türkiye
Welcome to the Age of Agent Ads: Using On-Policy Attacks to Steer Agentic Decisions
Oct 12, 2026Tim R. Davidson, Marija Šakota, Joshua Joseph +1Adversarial Attacks AI Agent Security
EPFL · Harvard University
Robust Detection of Directional Adversarial Attacks in Deep Neural Networks for Radiological Imaging
Oct 11, 2026Tristan Malatyński, Joanna Jaworek-KorjakowskaAdversarial Example Detection Adversarial Attacks
AGH University of Kraków, Poland Center of Excellence in Artificial Intelligence
From Acoustic Sensing to Physical Actions: A Comprehensive Survey of Security Risks in Edge-Deployed Audio LLM-Based Agents
Oct 11, 2026Binhong Wang, Haidong Kang, Yanfeng Zhang +3LLM Agent Security Edge AI
Northeastern University, China · Department of Computer Science, Hong Kong Baptist University, Hong Kong · Institute of Intelligent Manufacturing, Nanjing Tech University, Nanjing, China +1
Across Viewpoints and Time: Trajectory-Generalizable Physical Attacks on World Models and Video Understanding Models
Oct 11, 2026Song Xia, Meiwen Ding, Chenqi Kong +2Adversarial Attacks Physical Adversarial Attacks
Nanyang Technological University · National University of Singapore · Pengcheng Lab
X-AgentRed: Red-Teaming Personal Agents in Evolving Multi-Service Ecosystems
Oct 9, 2026Yeonsung Jung, Zhiyuan Wang, Zifan Wang +5Automated Red Teaming AI Agent Security
Meta Reality Labs Research · Meta Superintelligence Labs · Meta +1
Speedbumps: Rejection Attacks on Speculative Decoding
Oct 7, 2026Adam Y. J. Jones, Yu Yuan, Sergio MaffeisAdversarial Attacks Speculative Decoding
Imperial College London London, UK

Oct 29, 2026J/K move · Enter open · S save

PASA-KV: Attacking Action-Facing Visual Routes in World Action Models

Zhongbo Wu, Jiakai Huang, Zheng Zhang, Siyu Xu+5

Shanghai Jiao Tong University · Shanghai Innovation Institute · ACE Robotics+2

Adversarial Attacks on World Models Adversarial Attacks Action-Conditioned World Models Attention-Based Adversarial Attacks

Open paper AskPDF arXiv

World action models (WAMs) increasingly couple internal visual or world representations with action generation, making their world-to-action pathways a critical yet poorly understood source of robustness. We investigate when perturbations to these internal representations actually become behaviorally consequential, and uncover a pronounced mismatch between representation perturbability and closed-loop vulnerability: attack effectiveness varies sharply across visual K/V routes, while larger representation displacement does not reliably imply stronger control failure. Our analysis attributes this mismatch to the action-conditioned readout, where keys determine where action queries read and values determine what content they receive. This motivates an action-facing view of vulnerability, in which the effect of a perturbation depends on how it is routed into action generation rather than on representation-space magnitude alone. We instantiate this view with Tail-AV, a short-probe signal for identifying vulnerable routes, and operationalize it in PASA-KV, which selects action-facing K/V targets before training a fresh task-shared attack. Experiments across multiple WAM systems and a physical robot validate readout-informed target selection, demonstrate strong closed-loop attacks, and reveal substantial system-level differences in resistance.