Unsafer in Many Turns: Benchmarking and Defending Multi-Turn Safety Risks in Tool-Using Agents

• URL: http://arxiv.org/abs/2602.13379v1
• Date: Fri, 13 Feb 2026 18:38:18 GMT
• Title: Unsafer in Many Turns: Benchmarking and Defending Multi-Turn Safety Risks in Tool-Using Agents
• Authors: Xu Li, Simon Yu, Minzhou Pan, Yiyou Sun, Bo Li, Dawn Song, Xue Lin, Weiyan Shi,
• Abstract summary: We propose a principled taxonomy that transforms single-turn harmful tasks into multi-turn attack sequences.<n>Using this taxonomy, we construct MT-AgentRisk, the first benchmark to evaluate multi-turn tool-using agent safety.<n>We propose ToolShield, a training-free, tool-agnostic, self-exploration defense.
• Score: 68.20752678837377
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: LLM-based agents are becoming increasingly capable, yet their safety lags behind. This creates a gap between what agents can do and should do. This gap widens as agents engage in multi-turn interactions and employ diverse tools, introducing new risks overlooked by existing benchmarks. To systematically scale safety testing into multi-turn, tool-realistic settings, we propose a principled taxonomy that transforms single-turn harmful tasks into multi-turn attack sequences. Using this taxonomy, we construct MT-AgentRisk (Multi-Turn Agent Risk Benchmark), the first benchmark to evaluate multi-turn tool-using agent safety. Our experiments reveal substantial safety degradation: the Attack Success Rate (ASR) increases by 16% on average across open and closed models in multi-turn settings. To close this gap, we propose ToolShield, a training-free, tool-agnostic, self-exploration defense: when encountering a new tool, the agent autonomously generates test cases, executes them to observe downstream effects, and distills safety experiences for deployment. Experiments show that ToolShield effectively reduces ASR by 30% on average in multi-turn interactions. Our code is available at https://github.com/CHATS-lab/ToolShield.

Related papers

• ToolSafe: Enhancing Tool Invocation Safety of LLM-based agents via Proactive Step-level Guardrail and Feedback [53.2744585868162]
  Monitoring step-level tool invocation behaviors in real time is critical for agent deployment.<n>We first construct TS-Bench, a novel benchmark for step-level tool invocation safety detection in LLM agents.<n>We then develop a guardrail model, TS-Guard, using multi-task reinforcement learning.
  arXiv  Detail & Related papers  (2026-01-15T07:54:32Z)
• STAC: When Innocent Tools Form Dangerous Chains to Jailbreak LLM Agents [38.755035623707656]
  This paper introduces Sequential Tool Attack Chaining (STAC), a novel multi-turn attack framework that exploits agent tool use.<n>We apply our framework to automatically generate and evaluate 483 STAC cases, featuring 1,352 sets of user-agent-environment interactions.<n>Our evaluations show that state-of-the-art LLM agents, including GPT-4.1, are highly vulnerable to STAC, with attack success rates (ASR) exceeding 90% in most cases.
  arXiv  Detail & Related papers  (2025-09-30T00:31:44Z)
• OpenAgentSafety: A Comprehensive Framework for Evaluating Real-World AI Agent Safety [58.201189860217724]
  We introduce OpenAgentSafety, a comprehensive framework for evaluating agent behavior across eight critical risk categories.<n>Unlike prior work, our framework evaluates agents that interact with real tools, including web browsers, code execution environments, file systems, bash shells, and messaging platforms.<n>It combines rule-based analysis with LLM-as-judge assessments to detect both overt and subtle unsafe behaviors.
  arXiv  Detail & Related papers  (2025-07-08T16:18:54Z)
• Kaleidoscopic Teaming in Multi Agent Simulations [75.47388708240042]
  We argue that existing red teaming or safety evaluation frameworks fall short in evaluating safety risks in complex behaviors, thought processes and actions taken by agents.<n>We introduce new in-context optimization techniques that can be used in our kaleidoscopic teaming framework to generate better scenarios for safety analysis.<n>We present appropriate metrics that can be used along with our framework to measure safety of agents.
  arXiv  Detail & Related papers  (2025-06-20T23:37:17Z)
• SafeAgent: Safeguarding LLM Agents via an Automated Risk Simulator [77.86600052899156]
  Large Language Model (LLM)-based agents are increasingly deployed in real-world applications.<n>We propose AutoSafe, the first framework that systematically enhances agent safety through fully automated synthetic data generation.<n>We show that AutoSafe boosts safety scores by 45% on average and achieves a 28.91% improvement on real-world tasks.
  arXiv  Detail & Related papers  (2025-05-23T10:56:06Z)
• Think Twice Before You Act: Enhancing Agent Behavioral Safety with Thought Correction [20.7755316818041]
  Internal reasoning process significantly influences tool usage and subsequent actions.<n>Minor deviations in the agent's thought may trigger cascading effects leading to irreversible safety incidents.<n>Thought-Aligner corrects each high-risk thought on the fly before each action execution.<n>The corrected thought is then reintroduced to the agent, ensuring safer subsequent decisions and tool interactions.
  arXiv  Detail & Related papers  (2025-05-16T10:00:15Z)
• AgentBreeder: Mitigating the AI Safety Risks of Multi-Agent Scaffolds via Self-Improvement [5.5068092331643035]
  Scaffolding Large Language Models (LLMs) into multi-agent systems often improves performance on complex tasks.<n>We introduce AgentBreeder, a framework for multi-objective self-improving evolutionary search over scaffolds.
  arXiv  Detail & Related papers  (2025-02-02T11:40:07Z)
• SafeAgentBench: A Benchmark for Safe Task Planning of Embodied LLM Agents [58.65256663334316]
  We present SafeAgentBench -- the first benchmark for safety-aware task planning of embodied LLM agents in interactive simulation environments.<n>SafeAgentBench includes: (1) an executable, diverse, and high-quality dataset of 750 tasks, rigorously curated to cover 10 potential hazards and 3 task types; (2) SafeAgentEnv, a universal embodied environment with a low-level controller, supporting multi-agent execution with 17 high-level actions for 9 state-of-the-art baselines; and (3) reliable evaluation methods from both execution and semantic perspectives.
  arXiv  Detail & Related papers  (2024-12-17T18:55:58Z)
• Breaking Agents: Compromising Autonomous LLM Agents Through Malfunction Amplification [35.16099878559559]
  Large language models (LLMs) have experienced significant development and are being deployed in real-world applications. We introduce a new type of attack that causes malfunctions by misleading the agent into executing repetitive or irrelevant actions. Our experiments reveal that these attacks can induce failure rates exceeding 80% in multiple scenarios.
  arXiv  Detail & Related papers  (2024-07-30T14:35:31Z)
• Dissecting Adversarial Robustness of Multimodal LM Agents [70.2077308846307]
  We manually create 200 targeted adversarial tasks and evaluation scripts in a realistic threat model on top of VisualWebArena.<n>We find that we can successfully break latest agents that use black-box frontier LMs, including those that perform reflection and tree search.<n>We also use ARE to rigorously evaluate how the robustness changes as new components are added.
  arXiv  Detail & Related papers  (2024-06-18T17:32:48Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.