Skip to content

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

Oct 12 · Monday

All papers

Found 34 papers
Filter5

Attack type

Jailbreaks10Prompt injection20Detection evasion6

Attack technique

Automated search4Defense-targeted6Dormant triggers3

Defense type

Detection and filtering21Permissions and approval3

Impact

Harmful content9Unauthorized actions11Output manipulation9

System type

LLM applications13AI Agent23science agent3web agent3Multi-agent systems5

Attacker access

Black-box6

Attack surface / entry point

User input27Web pages and interfaces7Documents and messages10Tool responses10Tool and skill descriptions10Code repositories and dependencies4Retrieval corpus4Memory writes3Inter-agent messages3Images, audio and video6Training data0Distribution channels5

Layer

Prompt layer50Model layer3Application layer34Training and data layer3

System component

MCP and skills21Memory26RAG19Guardrails and judges34Monitors18Reasoning chains5Permissions and sandboxing29Vision12Multimodal13
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 233papers have no tags yet and are not included in filtered results.
Attack surface / entry point：User inputAttack surface / entry point：Web pages and interfacesAttack surface / entry point：Training dataLayer：Application layerSystem component：Guardrails and judgesClear all

Another 233papers do not have category tags yet and are not included in filtered results.

  1. AttacksarXiv 2610.12292

    The paper proposes an option channel attack: changing one option label can make an agent guardrail allow a prohibited operation

    Proposing an option channel attack that renames option labels so that agent guardrails allow prohibited operations

    arXiv
    2610.12292
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Attacker
    Black-box
    Evaluated models
    LAYA-TD, LAYA-EN, LAYA-ML and others5

    AbstractThe authors' evaluation indicates that typed models should not serve as final safety arbiters; they are susceptible to off-policy text and option name manipulation, and confidence filtering cannot resist attacks.

    Paper analysisRead analysis
  3. DefensesarXiv 2610.07657

    DEFER1 多智能体防御研究报告受控测试中的攻击成功率下降

    实现 DEFER，以确定性检查优先拦截多智能体流水线的内容攻击

    arXiv
    2610.07657
    Published
    Layer
    Application layer
    Setting
    Multi-agent systems
    Evaluated models
    Qwen3-235B-A22B-Instruct-2507, gpt-oss-120b, Llama-4-Scout-17B-16E-Instruct and others5
    AbstractDEFER 把确定性检查放在 judge 之前。

    DEFER 用来测量多智能体授权流水线里，确定性规则在何处结束、LLM judge 从何处开始。

    Paper analysisRead analysis
  4. AttacksarXiv 2610.05943

    CRIME 框架利用良性 Agent 技能组合诱发恶意行为

    提出 CRIME 框架，组合良性 Agent 技能诱发恶意行为

    arXiv
    2610.05943
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    DeepSeek-V4-Flash, GPT-4o-mini, GPT-4o and others6
    Abstract良性技能经 DCA 或 ACA 组合后可产生恶意环境后果，单独审查对不上。

    CRIME 把公共技能库中各自通过审查的技能配成两两组合，看联合执行能否实现指定恶意任务。。

    Paper analysisRead analysis
  5. DefensesarXiv 2610.04412

    论文研究 LLM 数字孪生参与增强民主时的提示注入脆弱性

    提出检测、结构化与 GSPO 流程，降低共识生成的提示注入错位

    arXiv
    2610.04412
    Published
    Layer
    Application layer
    Setting
    LLM applications, Multi-agent systems
    Evaluated models
    LLaMA-2 7B, LLaMA-3 8B, GPT 3.5 Turbo and others14
    Abstract三章依次做个体偏好预测、政党知识图谱审议，以及共识生成对提示注入的脆弱性与一条防御流程。

    这是一篇把 LLM 数字孪生放进增强民主的博士论文，分个体表示、政党聚合和共识生成的提示注入三章。

    Paper analysisRead analysis
  6. DefensesarXiv 2610.03434

    Persona Guardrail：面向智能体系统的生产级防御框架

    提出 Persona Guardrail，强制客服智能体守住功能边界

    arXiv
    2610.03434
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Attacker
    Black-box
    Evaluated models
    Qwen3-30B-A3B, Qwen3-VL-30B, Qwen3-8B and others5
    AbstractPersona Guardrail 用功能边界做同步输入输出校验，并在 PAGE 与生产延迟预算下给出评测。

    Persona Guardrail 是客服向智能体的同步运行时防御：先写明功能边界，再决定放行或拦截，而不是只检测某一攻击类。

    Paper analysisRead analysis
  7. DefensesarXiv 2610.02432

    论文提出 LLM 输入扰动鲁棒性度量，并发布 ASA 攻击与 AttestMCP 防护

    提出 ASA 与 AttestMCP，度量并降低 LLM 对输入序列扰动的敏感

    arXiv
    2610.02432
    Published
    Layer
    Application layer
    Setting
    LLM applications, AI Agent
    Attacker
    Black-box
    Evaluated models
    Gemma-3-27B-Instruct, Gemma-3-4B-Instruct, Llama-3.2-3B-Instruct and others12
    Paper analysisRead analysis
  8. DefensesarXiv 2609.33039

    TACIT: detecting agent trajectory harm from LLM internal states

    提出 TACIT，从冻结内部状态检测智能体有害轨迹

    arXiv
    2609.33039
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    Qwen3-4B, Llama-3.1-8B, Qwen3-4B-Instruct and others4

    Abstract提出基于内部表征的智能体轨迹安全检测方法 TACIT，揭示工具风险表征特性并在多项基准上提升检测表现与效率。

    Paper analysisRead analysis
  10. AttacksarXiv 2609.22510

    特拉维夫大学提出爆炸性提示词：条件触发的间接提示注入可绕过九款生产 Agent

    提出条件触发的爆炸提示词，并设计摄入期检测器 DeFuse

    arXiv
    2609.22510
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    Llama-3.1-8B-Instruct, Llama-3.1-70B-Instruct, Qwen2.5-32B-Instruct and others15
    Abstract论文揭示了条件触发使提示注入绕过防御并在触发时执行工具，提出轻量检测器并在摄入期有效防范。

    论文针对大语言模型智能体面临的间接提示注入威胁，研究了利用条件句式实现时间分离的爆炸提示词。

    Paper analysisRead analysis
  11. Evaluation and benchmarksarXiv 2609.15017

    PIDS-Bench：在过度防御、混淆与分布偏移下评测提示注入检测器

    提出 PIDS-Bench，评测提示注入检测器在过度防御与分布偏移下的表现

    arXiv
    2609.15017
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Evaluated models
    TF-IDF + logistic regression, DistilBERT, DeBERTa-v3-FT and others7
    AbstractPIDS-Bench 显示高分布内 F1 仍会过度拦截外部来源的安全相关良性输入。

    PIDS-Bench 是一个冻结的检测器级基准，用来在固定阈值下同时看提示注入检测和良性误报。

    Paper analysisRead analysis
  12. Evaluation and benchmarksarXiv 2609.02316

    Counter-GEO-Bench：现有护栏难挡 GEO 虚假信息

    构建 COUNTER-GEO-BENCH，评测生成式搜索中的 GEO 虚假信息防御

    arXiv
    2609.02316
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Evaluated models
    Gemma-4-31B-IT, Qwen-3.5-35B-A3B, Llama-4-Scout-17B-16E and others8

    Abstract论文构建了针对 GEO 虚假信息的首个防御基准，揭示通用护栏的失效与反作用，并提出了有效的轻量对比检测基线。

    Paper analysisRead analysis
  13. Evaluation and benchmarksarXiv 2609.05535

    Mind2Web-Injection：面向网页 Agent 视觉提示注入护栏的证据对齐评测

    提出 Mind2Web-Injection 评测网页 Agent 视觉注入护栏

    arXiv
    2609.05535
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Evaluated models
    GPT-5.6-luna, Qwen3-VL-32B, Qwen3-VL-8B and others6
    Abstract作者称检测、定位和反事实响应在六个 VLM 上可分开。

    Mind2Web-Injection 用来检查网页截图护栏的判决是否对齐到该用的画面证据。。

    Paper analysisRead analysis
  15. Evaluation and benchmarksarXiv 2608.16824

    GEO-Flag：检测与测量面向生成式引擎优化的网页内容

    提出 GEOFlagBench 与 IPT，检测并测量生成式引擎优化网页

    arXiv
    2608.16824
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Evaluated models
    ModernBERT-base, Qwen3-0.6B, Gemini-3.5-flash and others11
    AbstractGEOFlagBench 与 IPT 用于检测 GEO 页，并估计真实检索中的占比与引用可核验性。

    GEO-Flag 把页面级 GEO 检测做成可评测任务，并在发布的检索结果上估计被判为 GEO 的页面占比。

    Paper analysisRead analysis
  16. AttacksarXiv 2608.16246

    CompoSkill 框架利用逐个通过扫描的 Agent 技能组合发起攻击

    提出 CompoSkill，将逐个通过扫描的技能编排为攻击链

    arXiv
    2608.16246
    Published
    Layer
    Application layer
    Setting
    AI Agent
    Evaluated models
    GPT-5.4, Gemini-3.1-flash, DeepSeek-V4 and others4

    Abstract论文分析了智能体技能组合风险，指出单包安全认证状态不具备可组合性，需转向路径级防护。

    Paper analysisRead analysis
  19. DefensesarXiv 2608.12361

    基于搜索增强 LLM 的中文新词毒性共识检测框架 SeTox

    提出 SeTox 检索增强框架，检测中文新词的隐性毒性

    arXiv
    2608.12361
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Evaluated models
    SeTox-Qwen2.5-7B
    AbstractSeTox 用检索把公共共识接入静态 LLM，以检测中文毒性新词。

    SeTox 研究中文新词上的隐性毒性检测：词表加检索增强，使静态 LLM 在不重训练的情况下利用公开网页上下文。

    Paper analysisRead analysis
Load more