mlsanityData to 7 Oct 2026

Towards Robust and Secure Embodied AI: A Survey on Vulnerabilities and Attacks

arXiv 2502.13175 ↗First posted 18 Feb 2025cs.CR cs.AI cs.ROcited by 32

Not on the map. Why?

Abstract (arXiv)

Embodied AI systems, including robots and autonomous vehicles, are increasingly integrated into real-world applications, where they encounter a range of vulnerabilities stemming from both environmental and system-level factors. These vulnerabilities manifest through sensor spoofing, adversarial attacks, and failures in task and motion planning, posing significant challenges to robustness and safety. Despite the growing body of research, existing reviews rarely focus specifically on the unique safety and security challenges of embodied AI systems. Most prior work either addresses general AI vulnerabilities or focuses on isolated aspects, lacking a dedicated and unified framework tailored to embodied AI. This survey fills this critical gap by: (1) categorizing vulnerabilities specific to embodied AI into exogenous (e.g., physical attacks, cybersecurity threats) and endogenous (e.g., sensor failures, software flaws) origins; (2) systematically analyzing adversarial attack paradigms unique to embodied AI, with a focus on their impact on perception, decision-making, and embodied interaction; (3) investigating attack vectors targeting large vision-language models (LVLMs) and large language models (LLMs) within embodied systems, such as jailbreak attacks and instruction misinterpretation; (4) evaluating robustness challenges in algorithms for embodied perception, decision-making, and task planning; and (5) proposing targeted strategies to enhance the safety and reliability of embodied AI systems. By integrating these dimensions, we provide a comprehensive framework for understanding the interplay between vulnerabilities and safety in embodied AI.

Key facts. Cited by 32 papers in the map (data through 7 Oct 2026). Builds on 141 of its 269 references that are in the map. First posted on arXiv 18 Feb 2025.

Builds on

141 of 269 references.

References · 141

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  2. 2020Language Models are Few-Shot Learnerscited by 32,199
  3. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 7,916
  4. 2023Visual Instruction Tuningcited by 7,792
  5. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  6. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  7. 2013Intriguing properties of neural networkscited by 5,114
  8. 2022Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedbackcited by 3,634
  9. 2023Generative Agents: Interactive Simulacra of Human Behaviorcited by 3,500
  10. 2016Towards Evaluating the Robustness of Neural Networkscited by 2,976
  11. 2021TruthfulQA: Measuring How Models Mimic Human Falsehoodscited by 2,919
  12. 2023Universal and Transferable Adversarial Attacks on Aligned Language Modelscited by 2,914
38 more
  1. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,777
  2. 2016Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answeringcited by 2,498
  3. 2022Do As I Can, Not As I Say: Grounding Language in Robotic Affordancescited by 2,302
  4. 2024π0: A Vision-Language-Action Flow Model for General Robot Controlcited by 2,297
  5. 2020Extracting Training Data from Large Language Modelscited by 2,144
  6. 2022RT-1: Robotics Transformer for Real-World Control at Scalecited by 2,007
  7. 2023ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIscited by 1,880
  8. 2015Microsoft COCO Captions: Data Collection and Evaluation Servercited by 1,447
  9. 2015Flickr30k Entities: Collecting Region-to-Phrase Correspondences for Richer Image-to-Sentence Modelscited by 1,352
  10. 2024HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusalcited by 1,330
  11. 2023Jailbreaking Black Box Large Language Models in Twenty Queriescited by 1,326
  12. 2020RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Modelscited by 1,235
  13. 2022Code as Policies: Language Model Programs for Embodied Controlcited by 1,172
  14. 2019Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learningcited by 1,168
  15. 2023Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversationscited by 1,124
  16. 2019OK-VQA: A Visual Question Answering Benchmark Requiring External Knowledgecited by 1,114
  17. 2023Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!cited by 1,106
  18. 2016Modeling Context in Referring Expressionscited by 1,054
  19. 2017Vision-and-Language Navigation: Interpreting visually-grounded navigation instructions in real environmentscited by 1,013
  20. 2020StereoSet: Measuring stereotypical bias in pretrained language modelscited by 911
  21. 2015Generation and Comprehension of Unambiguous Object Descriptionscited by 897
  22. 2023BeaverTails: Towards Improved Safety Alignment of LLM via a Human-Preference Datasetcited by 828
  23. 2015The Limitations of Deep Learning in Adversarial Settingscited by 784
  24. 2014Deep Neural Networks are Easily Fooled: High Confidence Predictions for Unrecognizable Imagescited by 774
  25. 2019ALFRED: A Benchmark for Interpreting Grounded Instructions for Everyday Taskscited by 773
  26. 2023AutoDAN: Generating Stealthy Jailbreak Prompts on Aligned Large Language Modelscited by 738
  27. 2022Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learnedcited by 702
  28. 2023Baseline Defenses for Adversarial Attacks Against Aligned Language Modelscited by 608
  29. 2022Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulationcited by 545
  30. 2019The Woman Worked as a Babysitter: On Biases in Language Generationcited by 540
  31. 2024Executable Code Actions Elicit Better LLM Agentscited by 537
  32. 2024How Johnny Can Persuade LLMs to Jailbreak Them: Rethinking Persuasion to Challenge AI Safety by Humanizing LLMscited by 533
  33. 2017Adversarial Patchcited by 514
  34. 2023"Do Anything Now": Characterizing and Evaluating In-The-Wild Jailbreak Prompts on Large Language Modelscited by 457
  35. 2016Transferability in Machine Learning: from Phenomena to Black-Box Attacks using Adversarial Samplescited by 439
  36. 2023GPT-4 Is Too Smart To Be Safe: Stealthy Chat with LLMs via Ciphercited by 401
  37. 2023Jailbreak and Guard Aligned Language Models with Only Few In-Context Demonstrationscited by 401
  38. 2021Persistent Anti-Muslim Bias in Large Language Modelscited by 376

50 most cited of 141 All 141 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 13; 2026: 19.

32 papers.

Sort:

By area:

  1. 2025IS-Bench: Evaluating Interactive Safety of VLM-Driven Embodied Agents in Daily Household Taskscited by 40
  2. 2025MCP-Guard: A Multi-Stage Defense-in-Depth Framework for Securing Model Context Protocol in Agentic AIcited by 19
  3. 2025A Survey on Agentic Security: Applications, Threats and Defensescited by 17
  4. 2025Goal-oriented Backdoor Attack against Vision-Language-Action Models via Physical Objectscited by 16
  5. 2025Self-Destructive Language Modelcited by 10
  6. 2025Reasoning or Retrieval? A Study of Answer Attribution on Large Reasoning Modelscited by 9
  7. 2026Unveiling the Fragility of Vision-Language Models: Multi-Modal Adversarial Synergy via Texture-Constrained Perturbations and Cross-Modal Optimizationcited by 9
  8. 2026Inject Once Survive Later: Backdooring Vision-Language-Action Models to Persist Through Downstream Fine-tuningcited by 7
  9. 2025Embracing Trustworthy Brain-Agent Collaboration as Paradigm Extension for Intelligent Assistive Technologiescited by 5
  10. 2025A Review of 3D Object Detection with Vision-Language Modelscited by 4
  11. 2025Embodied AI: Emerging Risks and Opportunities for Policy Actioncited by 3
  12. 2026LABSHIELD: A Multimodal Benchmark for Safety-Critical Reasoning and Planning in Scientific Laboratoriescited by 3
20 more
  1. 2026Tri-Info: Generalizable, Interpretable Failure Prediction for VLA Models via Information Theorycited by 3
  2. 2025ANNIE: Be Careful of Your Robotscited by 2
  3. 2026Security of World-Model-Based Embodied AI: A Lifecycle of Threats, Defenses, and Evaluationcited by 2
  4. 2026HomeSafe-Bench: Evaluating Vision-Language Models on Unsafe Action Detection for Embodied Agents in Household Scenarioscited by 1
  5. 2026CMP: Robust Whole-Body Tracking for Loco-Manipulation via Competence Manifold Projectioncited by 1
  6. 2026RIPA: Sensory-Vector Prompt Injection Attacks on LLM-Controlled ROS 2 Robotscited by 1
  7. 2025Large Language Models in Cybersecurity: Applications, Vulnerabilities, and Defense Techniquescited by 0
  8. 2025HGMF: A Hierarchical Gaussian Mixture Framework for Scalable Tool Invocation within the Model Context Protocolcited by 0
  9. 2025Fundamentals of Physical AIcited by 0
  10. 2026Silencing the Guardrails: Inference-Time Jailbreaking via Dynamic Contextual Representation Ablationcited by 0
  11. 2026RedVLA: Physical Red Teaming for Vision-Language-Action Modelscited by 0
  12. 2026NEXUS: Continual Learning of Symbolic Constraints for Safe and Robust Embodied Planningcited by 0
  13. 2026When Agents Control Robots: A Zero Trust Policy Model for Agentic Cyber-Physical Systemscited by 0
  14. 2026Mastermind: Strategy-grounded Learning for Repository-Scale Vulnerability Reproductioncited by 0
  15. 2026AdvNav: Behavior-Guided Black-Box Adversarial Attacks on Vision-Language Navigationcited by 0
  16. 2026Self-Evolving Just-In-Time Memory for Proactive Embodied Safetycited by 0
  17. 2026On Understanding, Identifying, and Mitigating Vulnerabilities in Agentic Large Language Modelscited by 0
  18. 2026FailureSpot: Label-Efficient Timestamp-Level Failure Detection for Vision-Language-Action Modelscited by 0
  19. 2026A Comprehensive Review of Generative Physical Artificial Intelligencecited by 0
  20. 2026Security and Privacy in Large-Model-Driven Embodied Agents: Attacks, Defenses, and Future Directionscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Safety in Embodied AI: A Survey of Risks, Attacks, and Defenses
  2. 2026RoboJailBench: Benchmarking Adversarial Attacks and Defenses in Embodied Robotic Agents
  3. 2026Security of Foundation-Model-Powered Embodied Agents: Attack Surfaces, Attacks, Defenses, and Evaluation
  4. 2025CHAI: Command Hijacking against embodied AI
  5. 2024Can We Trust Embodied Agents? Exploring Backdoor Attacks against Embodied LLM-based Decision-Making Systems
  6. 2026Silent Failures in Physical AI: A Literature Review of Runtime Action Authorization for Autonomous Systems
  7. 2025AGENTSAFE: Benchmarking the Safety of Embodied Agents on Hazardous Instructions
  8. 2025Advancing Embodied Agent Security: From Safety Benchmarks to Input Moderation