Skip to content

Translation in progress Some entries still include text in the original language.

Oct 13 · Tuesday

All papers

Found 23 papers

Another 201papers do not have category tags yet and are not included in filtered results.

  1. DefensesarXiv 2610.09793

    用时序逻辑监控工具 Agent

    提出 MFOTL 离线时序验证以监控工具 Agent 的危险动作链

    Published
    Setting
    AI Agent
    Attacker
    Black-box
    Abstract形式化验证可有效识别多步攻击,但其有效性高度依赖轨迹的可信历史记录。

    这篇论文对工具调用大语言模型智能体的安全监控进行了形式化时序验证研究。

    Paper analysisRead analysis
  2. DefensesarXiv 2610.05163

    Researchers construct segmented prompt injection on Claude Code and Codex and release a boundary action audit benchmark

    提出 PAA 方法,在边界动作前审计长流程智能体的分阶段提示注入

    Published
    Evaluated models
    Claude Code (v2.1.172, Claude Sonnet 5), Codex CLI (v0.144.0, GPT-5.5)
    Abstract论文针对长流程注入提出边界动作审计与 PAA,在保持低误阻的同时实现高拦截率。

    本研究针对长流程智能体在执行任务时面临的分阶段间接提示注入威胁,提出了在动作生效前进行拦截的边界动作审计框架与路径对齐归因方法。

    Paper analysisRead analysis
  3. MADBench: multi-agent debate safety evaluation benchmark released

    提出 MADBench,评测多智能体辩论是吸收还是放大对抗影响

    Published
    Evaluated models
    gpt-4o, GPT-5, GPT-6 Astra
    AbstractMADBench 显示辩论能吸收部分准确率攻击,同时放大未授权读写。

    MADBench 是一个评测多智能体辩论安全性的基准,用来看攻击是被辩论吸收还是被放大。

    Paper analysisRead analysis
  4. DefensesarXiv 2609.35659

    Tracekit: A tamper-proof intent-reasoning-action audit system for autonomous coding agents

    提出 Tracekit,审计编码 Agent 的意图、推理与动作并检测账本篡改

    Published
    Evaluated models
    Claude Code (claude -p, version 2.1), 独立评审(同一 CLI,无工具)
    AbstractTracekit 把意图、自述和动作写入可锚定的哈希链。

    Tracekit 是无第三方依赖的编码智能体审计层:把人类请求、模型自述和已执行动作并排写入可外部锚定的哈希链,执行前用正则门,事后用规则和独立模型互查。

    Paper analysisRead analysis
  5. 研究显示 Claude Code、Codex 等本地智能体可轻易篡改自身执行轨迹

    测试编程智能体被诱导或自主篡改自身执行轨迹的行为

    Published
    Evaluated models
    GPT-5.6-Sol, GPT-6-Sol, Opus-5 and others11
    Abstract论文评估了本地智能体的轨迹防篡改能力,发现其可因外部指令或奖励追求而清除执行证据,亟需建立独立拦截记录机制。

    这是一项针对本地大模型智能体执行轨迹完整性与安全防篡改边界的实证研究。

    Paper analysisRead analysis
  6. 研究提出 ACE 语料库,评估内核级证据对 Agent 安全检测的有效性

    构建 ACE 语料,评测内核级证据对 Agent 攻击检测的有效性

    Published
    Evaluated models
    AdaBoost, XGBoost, TabPFN and others10
    AbstractACE 用配对的内核与应用层证据表明,syscall 痕迹可判别智能体攻击,两层组合通常更好。

    ACE 是一份把智能体会话的内核 syscall 痕迹与应用层清单、转录配对起来的检测语料和基准。。

    Paper analysisRead analysis
  7. AttacksarXiv 2609.27542

    Researchers use control token injection to delete gpt-oss-20b's reasoning chain and bypass chain-of-thought monitoring

    提出控制标记注入,清空工具 Agent 的思维链以绕过推理监控

    Published
    Setting
    AI Agent
    Evaluated models
    openai/gpt-oss-20b, google/gemma-4-26B-A4B-it, deepseek-ai/DeepSeek-R1-Distill-Qwen-7B and others4

    Abstract论文阐明控制标记注入可抑制 CoT 使监控失效并促成拒绝绕过,且工具执行受解析器宽容度影响。

    Paper analysisRead analysis
  8. AttacksarXiv 2609.12448

    GraphProfiler 用个人知识图谱推断敏感属性并溯源到具体帖子

    提出 GraphProfiler,用个人知识图谱从帖子历史推断敏感属性并溯源

    Published
    AbstractGraphProfiler 用源链接个人知识图谱做属性推断,准确率接近文本基线,并把证据定位到少量源帖。

    GraphProfiler 是一个可审计的 LLM 敏感属性画像器:用源链接个人知识图谱,把推断落到具体帖子。

    Paper analysisRead analysis
  9. AttacksarXiv 2609.02774

    CodePoisonRAG:针对检索增强代码生成的知识投毒攻击

    提出 CodePoisonRAG,对检索增强代码生成的知识库投毒以植入漏洞

    Published
    Evaluated models
    Qwen 3.5 9B, Code Llama 13B, DeepSeek-Coder-V2 16B

    AbstractCodePoisonRAG 展示了 RACG 外部知识库的新攻击面,在较低投毒率下可定向诱发指定漏洞并部分抵御防御。

    Paper analysisRead analysis
  10. DefensesarXiv 2608.21101

    ClawSentry:面向自主 LLM Agent 的渐进式多层安全监控网关

    提出 ClawSentry 多层监控网关,在技能准入和运行时拦截恶意技能与隐藏指令

    Published
    Evaluated models
    Codex/GPT-5.4, Codex/GPT-5.5, Claude Code/GLM-5.1 and others6
    AbstractClawSentry 用包准入、三级运行时门控和反绕过,压低 skill 攻击并保住干净任务。

    ClawSentry 是挂在智能体运行时之外的安全监督网关,用来同时盖住 skill 执行的多个生命周期边界,并把已被拒绝的效果记到会话里。

    Paper analysisRead analysis
  11. DefensesarXiv 2605.17380

    Uber proposes ADR enterprise-level agent detection system, deployed on 7,200 hosts for ten months

    提出企业级检测系统 ADR,识别编程 Agent 的提示注入与恶意 MCP

    Published

    Abstract提出了端点因果遥测、两级在线分流与离线红队闭环的 ADR 系统,并在生产环境中验证了有效性。

    Paper analysisRead analysis
  12. AttacksarXiv 2603.17239

    LAAF: A Systematic Red-Teaming Method and Automated Attack Framework for LPCI Vulnerabilities in Agentic LLM Systems

    Proposes LAAF, which automatically attacks LPCI vulnerabilities in agent systems in stages

    Published
    Setting
    AI Agent
    Evaluated models
    gemini-2.0-flash-001, claude-3-haiku-20240307, meta-llama/llama-3.1-70b-instruct and others5
    AbstractLAAF does LPCI red teaming with taxonomy plus stage seeding.

    The authors state that LAAF is an automated red-teaming framework built specifically for LPCI in agentic LLMs. The issue to address is that encoded, delayed, and conditionally triggered payloads can persist in memory or RAG and re-enter across sessions, whereas Garak and PyRIT do not use the previous round's breakthrough results to seed the next stage across the six stages.

    Paper analysisRead analysis
  13. DefensesarXiv 2602.20708

    ICON:基于推理时纠正的 Agent 间接提示注入防御

    Proposes inference-time defense ICON, which detects attention over-focusing and redirects to correct indirect prompt injection in agents.

    Published
    Setting
    AI Agent
    Evaluated models
    Qwen-3-8B, Llama-3.1-8B, Mistral-8B and others6
    AbstractICON detects over-focusing and redirects at inference time to balance IPI safety and task utility.

    ICON detects attention over-focusing from IPI at inference time and redirects, avoiding refusal that would terminate the agent trajectory.

    Paper analysisRead analysis
  14. AttacksarXiv 2602.15654

    Zombie Agent: Persistent Control of Self-Evolving LLM Agents via Self-Reinforcing Injection

    提出 Zombie Agent 攻击,经网页间接注入将载荷写入长期记忆并跨会话触发

    Published
    Setting
    AI Agent
    Evaluated models
    Gemini-2.5-Flash, GLM-4.7-Flash
    Abstract记忆写入把一次网页间接注入变成跨会话控制。

    Zombie Agent 把自进化 LLM 智能体的跨会话记忆当成攻击面:攻击者只发布网页,良性任务读入后,正常记忆更新留下载荷,并在以后无关会话里触发未授权工具行为。

    Paper analysisRead analysis
  15. When benchmarks lie: evaluating malicious prompt classifiers under real-world distribution shift

    提出 LODO 评测恶意提示探针的跨数据集泛化

    Published
    Evaluated models
    Llama-3.1-8B-Instruct, Gemma-3-27B, Qwen-3.5-4B and others5
    Abstract作者称 LODO 露出 CV 高估,数据集身份与安全方向部分重叠,标准域泛化干预关不上。

    LODO 用来检查恶意提示词分类器换成没见过的数据集后,还剩多少检测能力。。

    Paper analysisRead analysis