AI Agent Security Benchmarks

Ask about this topic

Overview

AI agent security benchmarks test whether agents remain safe as they use tools, retrieve information, manage memory, and act across multi-step workflows. Common approaches use sandboxed environments, realistic attack scenarios, adaptive red-teaming, and executable checks of tool calls or resulting system states; cyber ranges also assess vulnerability discovery, exploitation, and repair. The field is moving beyond single-turn prompt-injection tests to cover persistent memory, third-party skills, privacy, multi-agent interactions, and real-world interfaces. A central challenge is making results both realistic and comparable: static tests may overstate robustness, while attack success alone can miss harm severity, authorization failures, and the cost to legitimate task utility.

Latest papers

CardsList

309 papers56 papers in the last four weeks, up 51% on the four weeks before. 0.3% of all new papers.

  1. Judging Before Propagation: Task-Graph Contextual Integrity for Multi-Agent Privacy

    Oct 29, 2026Hengyu Wu, Blaise Delattre, Yang CaoPrivacy Protection for LLM AgentsContextual Integrity

    Department of Computer Science, School of Computing, Institute of Science Tokyo

  2. Where You Inject Matters: The Trust Region Effect in LLM Agents

    Oct 28, 2026Chengquan Guo, Xiaodong Liu, Weiwei Yang +7Prompt InjectionPrompt Injection Attacks on AI Agents

    University of Chicago · Microsoft Research · University of Rochester +3

  3. How Deterministic SAST and AI Security Agents Divide the Work: Coverage, Harness Policy and Cost in Snyk VulnBench 2.0

    Oct 28, 2026Liran Tal, Chris Liu, Charlotte Fereday +2Software Vulnerability DetectionAI Agent Security Benchmarks

    Snyk

  4. Scan Before You Install: Detecting and Mitigating Local and Composite Vulnerabilities in Skills

    Oct 27, 2026Parth A. Gandhi, Avital Aviv, Prasanna N. Wudali +2Agent Skill Security AuditingSoftware Vulnerability Detection

    Stein Faculty of Computer and Information Science, Ben-Gurion University of the Negev, Beer-Sheva, Israel

  5. Beyond QASecClaw: Agent Scaling, Verifier Feedback, and Evidence Reliability in Software Assurance

    Oct 27, 2026S M Abdulla Al Mamun, Akif Islam, Mohd. Ruhul AmeenLLM Agent VerificationLLM Reliability

    Department of Computer Science, The University of Manchester, United Kingdom · Department of Computer Science and Engineering, University of Rajshahi, Bangladesh · Department of Computer Science and Engineering, Varendra University, Bangladesh

  6. The Agent in the Middle: How AI-Mediated Access Reshapes Web Vulnerability Exploitation

    Oct 26, 2026Ruozhao Yang, Jiongchi Yu, Ming Hu +1AI Agent SecurityAI Agent Security Benchmarks

    Singapore Management University, Singapore · Nanyang Technological University, Singapore · East China Normal University, China

  7. SkillLifeBench: Benchmarking Security Across the Lifecycle of Agent Skills

    Oct 26, 2026Chenxu Zhang, Zhou Liu, Wu Xin +1AI Agent Security BenchmarksLLM Agent Evaluation

    Imperial College London, United Kingdom · Peking University, China

  8. CoSA: The Conflict of Skills Attack in Real-World LLM Agent Ecosystems

    Oct 26, 2026Yujia Zheng, Tianhao Li, Shangzhi Xu +6LLM Agent SecurityLLM Agent Safety

    University of Science and Technology Beijing · ZODA · Fangcun Leap +5

  9. Scope-Gap Injection: Exploiting Evaluator Channel Blind Spots in the OpenClaw Agent Security Competition

    Oct 26, 2026Weihang Niu, Mengxing Ren, Che QuPrompt Injection Attacks on AI AgentsIndirect Prompt Injection

    University of Science and Technology of China · University of Electronic Science and Technology of China

  10. CyberDual: AI Attackers versus AI Code Reviewers in Real Software Maintenance

    Oct 25, 2026Chenyuan Yang, Hwiwon Lee, Jongseong Kim +4AI Agent Security BenchmarksAutomated Code Review

  11. SEEK: Skill Extraction with Evolving Knowledge against Frontier Science Agents

    Oct 25, 2026Cheng Gui, Xiang Zheng, Yutao Wu +5AI Agent SecurityAgent Skill Security Auditing

    City University of Hong Kong, Hong Kong · Hong Kong Institute of AI for Science, Hong Kong · Deakin University, Australia +2

  12. Can Agentic Code Review Catch Vulnerabilities Before They Are Merged?

    Oct 24, 2026Yihan Yang, Chenyuan Yang, Zijie Zhao +5AI Agent Security BenchmarksAutomated Code Review

    University of Illinois Urbana-Champaign · Quantum AI & Cyber Security Institute, FPT Corporation; work done during an internship at UIUC.

  13. Agent-to-Agent Synthetic Social Engineering on the Web

    Oct 24, 2026Bryan C. BootsAI Agent SecurityWeb Agents

    University of Missouri-Kansas City, Kansas City, MO USA

  14. A Deterministic Security Runtime for LLM Agents on MPI and Batch-Scheduled HPC Systems: Consensus-Governed, Hardware-Rooted Supervision of Every Rank

    Oct 23, 2026Alfredo MetereLLM Agent SecurityRuntime Enforcement for AI Agents

    Enclawed, Inc., California, USA

  15. Beyond Copying: Lifelong Smart Contract Security Task Construction via Protocol Mutation

    Oct 23, 2026Zifan Peng, Peng Huang, Jingyi Zheng +5Software SecuritySynthetic Benchmark Generation

    Shanghai Artificial Intelligence Laboratory · The Hong Kong University of Science and Technology (Guangzhou) · University of Science and Technology of China +1

  16. AgentPhishing: A Benchmark of LLM Email Agents Under Adaptive Attack

    Oct 22, 2026Justin Tan, Miao Li, Ding ZhaoLLM Agent SecurityAI Agent Security Benchmarks

    Carnegie Mellon University

  17. AgentPolicyForge: From Executable Environments to Policy Benchmarks and Adaptive Guards

    Oct 20, 2026Chao Shang, Wanyu Du, Yi Zhang +1Tool Access Control for LLM AgentsLLM Guardrails

    AWS AI Labs

  18. Attention-based Classification for Indirect Prompt Injection

    Oct 20, 2026Amelia Kawasaki, Stefan LeePrompt Injection DetectionPrompt Injection Attacks on AI Agents

    Corvallis, Oregon, USA

  19. After the Attacker Leaves: Benchmarking Carryover Attacks on LLM Agents

    Oct 20, 2026Zhiyue Xu, Xukai Wang, Zhiyuan Huang +3Adversarial Attacks on LLMsPrompt Injection Attacks on AI Agents

    MAIS, Institute of Automation, Chinese Academy of Sciences · School of Artificial Intelligence, UCAS · Zhongguancun Academy +2

  20. The Show-Up Attack: When Subagents Win Over Their Supervisor

    Oct 20, 2026Maosen Zhang, Jianshuo Dong, Yutong Zhang +3LLM Agent SecurityMulti-Agent System Security

    Tsinghua University · SiliconProspect AI · Nanyang Technological University

  21. AutoRedClaw: Automated Red-Teaming for Claw-Style Agents

    Oct 19, 2026Hao Li, Tongyan Hu, Weidi Luo +8AI Agent SecurityAutomated Red Teaming

    Washington University in St. Louis · Johns Hopkins University · National University of Singapore +3

  22. PROVISO: Provenance-Informed Action Selection for Secure LLM Agents

    Oct 19, 2026Bo Yan, Qi Zhang, Ziwen Kan +2LLM Agent SecurityLLM Agents

    University of Central Florida · Independent Researcher · University of Tennessee, Knoxville

  23. Unbounded Losses, Unchecked Inputs: Measuring the Safeguards of LLM Options-Trading Agents

    Oct 16, 2026Mengxiao Wang, Nitesh SaxenaLLM Agent SecurityAI Agent Security Benchmarks

    Texas A&M University

  24. SPIDES: Information-Flow Control for Browser Agents

    Oct 16, 2026Egor Zverev, Juhee Kim, Leon Trampert +4LLM Agent SecurityInformation Flow Control

    ISTA · Microsoft · CISPA Helmholtz Center for Information Security

  25. Red-Teaming Multi-Agent Attacks for Loss-of-Control Risks

    Oct 16, 2026John Kitaoka, Thao Amelia Pham, Carlos Rafael Giudice +1AI Agent SafetyMulti-Agent System Security

    MATS Research · Redwood Research · Independent

  26. GatewayBench: Unified Capability and Security Evaluation Platform for MCP Gateways

    Oct 16, 2026Yitian Zheng, Zhaorun Chen, Haoyu Li +1Model Context ProtocolAI Agent Security Benchmarks

    Tsinghua University · University of Chicago · University of Illinois Urbana-Champaign

  27. MobileInject: Evaluating Reasoning-to-Action Security of MLLM-Based Mobile Agents under Cross-Modal Consistency Injection

    Oct 16, 2026Songping Wang, Peiyang Li, Yi Huang +5LLM Agent SecurityAI Agent Security Benchmarks

  28. INFUSE: A Unified Surface Benchmark for Indirect Prompt Injection in Agentic Systems

    Oct 14, 2026Thuy Dung Nguyen, Jingyi Guo, Rana Seda Dubauskas +3Indirect Prompt InjectionPrompt Injection Attacks on AI Agents

    Vanderbilt University

  29. CoRedSkill: Stealthy Skill Composition Attacks via AST-Guided Injection

    Oct 14, 2026Qichen Zhao, Shengfang Zhai, Qingni Shen +2AI Agent SecurityAutomated Red Teaming

    Peking University · National University of Singapore

  30. AgentCIScope: Grounded Contextual Integrity Evaluation for LLM Agents

    Oct 13, 2026Chengquan Guo, Zhaorun Chen, Zinan Lin +2Privacy Protection for LLM AgentsContextual Integrity

    University of Chicago · Microsoft Research · University of Illinois Urbana-Champaign