Skip to content

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

Oct 12 · Monday

All papers

Found 43 papers
Filter5

Attack type

Prompt injection19Poisoning and backdoors6Extraction and theft8Detection evasion8

Attack technique

Automated search3Role-play and persuasion3

Defense type

Model hardening3Detection and filtering8Instruction–data separation5Permissions and approval3

Impact

Harmful content5Unauthorized actions8Data and privacy leaks9Output manipulation19

System type

LLM applications34AI Agent158coding agent19science agent5web agent26Browser and computer use9Multi-agent systems13Embodied systems and robotics3

Attacker access

Black-box12

Attack surface / entry point

User input29Web pages and interfaces14Documents and messages20Tool responses5Code repositories and dependencies3Retrieval corpus24Images, audio and video10

Layer

Prompt layer4Application layer43

System component

RAG14Guardrails and judges13Vision5Multimodal5
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.
System type：LLM applicationsSystem type：Browser and computer useAttack surface / entry point：User inputAttack surface / entry point：Web pages and interfacesLayer：Application layerClear 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. AttacksarXiv 2610.11932

    A Fudan team measures the citation fragility of AI search: ordinary posts can enter citations and answer text

    Measures citation fragility in AI search, demonstrating that ordinary public posts can enter citations and answers.

    arXiv
    2610.11932
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Evaluated models
    Grok, Doubao, ChatGPT and others10

    AbstractThe authors claim the paper quantifies vulnerabilities in AI search caused by citation preferences and low posting thresholds; tests show public web posts can penetrate AI citations, and they call for auditing the citation supply chain.

    Paper analysisRead analysis
  5. DefensesarXiv 2610.07276

    SAFESHIELD：小语言模型部署时安全的决策组织框架

    提出 SAFESHIELD，组织小语言模型部署时的安全决策

    arXiv
    2610.07276
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Evaluated models
    Qwen3-4B-Instruct
    AbstractSAFESHIELD 用可审计决策轨迹组织 SLM 部署时的四项安全决策，并报告协调依赖影响端到端结果。

    SAFESHIELD 是面向小语言模型的部署时安全框架，把运行时安全从单项审核改写成可检查的决策过程。

    Paper analysisRead analysis
  6. AttacksarXiv 2610.04985

    研究揭示 Jev 作为应用决策层的安全与隐私风险

    评估应用决策层的输入注入与隐私推断风险

    arXiv
    2610.04985
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Attacker
    Black-box, Gray-box
    Evaluated models
    Jev, NanoJev
    AbstractJev 的类型化决策可被输入与训练操纵，也会泄露信息，并同时支持检测与滥用。

    Jev 是 TypeSafe 的 System One 模型，把应用 state 和固定问题变成 Choice、Noul 或 Score 的类型化决策与概率。作者用官方 API 和可改训练的 NanoJev，考察它作为决策层时的操纵、泄露与双重用途。

    Paper analysisRead analysis
  7. 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
  8. 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
  9. DefensesarXiv 2610.03089

    COBRA: A Dual-LLM Architecture for Defending Computer-Using Agents Against Branch-Jumping Attacks

    提出 COBRA 双 LLM 架构，防御计算机使用 Agent 的分支转向攻击

    arXiv
    2610.03089
    Published
    Layer
    Application layer
    Setting
    Browser and computer use
    Evaluated models
    gpt-5, UI-TARS-1.5-7B, claude-sonnet-5 and others6

    Abstract提出分支转向防御架构 COBRA，通过事前能力约束与边界代理消除分支转向攻击并兼顾效用。

    Paper analysisRead analysis
  12. Evaluation and benchmarksarXiv 2610.06898

    DIBench：面向移动 GUI 智能体多候选选择决策完整性的安全基准

    用 DIBench 评测移动 GUI 智能体被 UI 注入后的决策完整性

    arXiv
    2610.06898
    Published
    Layer
    Application layer
    Setting
    Browser and computer use
    Evaluated models
    Qwen2.5-VL(32B), Qwen3-VL(32B), GUI-Owl(32B) and others7

    Abstract论文提出移动智能体决策完整性基准 DIBench，发现欺诈 UI 注入诱导目标选择并推高完成率，通用防御收益不稳定。

    Paper analysisRead analysis
  13. DefensesarXiv 2609.36326

    PILLAR：面向增强检索的私有倒排索引词法查找

    提出 PILLAR，用私有词法预筛与客户端重排保护 RAG 查询

    arXiv
    2609.36326
    Published
    Layer
    Application layer
    Setting
    LLM applications
    AbstractPILLAR 用固定模式的私有 BM25 预筛加本地嵌入重排，在两基准上换取延迟或质量。

    PILLAR 是面向公开语料的隐私 RAG：半诚实服务器学不到查询词项和访问模式，客户端仍拿到用于增强生成的 k 篇文档。

    Paper analysisRead analysis
  15. DefensesarXiv 2609.22724

    MATE：面向移动 Agent 的策略感知安全审计方法

    提出移动 Agent 轨迹审计器 MATE，判断轨迹是否违反可编辑安全策略

    arXiv
    2609.22724
    Published
    Layer
    Application layer
    Setting
    Browser and computer use
    Attacker
    Black-box
    Evaluated models
    MATE-0.5B, MATE-1.5B, MATE-3B
    AbstractMATE 用合成轨迹微调轻量审计器，在可编辑策略下判断移动智能体轨迹是否违规并给出解释。

    MATE 是面向移动 GUI 智能体的轻量策略条件轨迹审计器，在给定自然语言安全策略时判断轨迹是否违规并解释原因。

    Paper analysisRead analysis
  17. 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
  18. AttacksarXiv 2609.14649

    CIG-MIA：利用上下文信息增益对 RAG 系统发起成员推断攻击

    提出 CIG-MIA，用上下文信息增益推断 RAG 知识库成员身份

    arXiv
    2609.14649
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Attacker
    Gray-box, Black-box
    Evaluated models
    Llama-3.1-8B-Instruct, Mistral-7B-Instruct-v0.3, Llama-2-13b-chat-hf and others6
    AbstractCIG-MIA 用注入前后的答案似然差推断 RAG 库成员，灰盒可分性高于所列基线。

    CIG-MIA 是针对 RAG 外部知识库的成员推断攻击，同时覆盖灰盒和纯文本黑盒。

    Paper analysisRead analysis
  19. DefensesarXiv 2609.10992

    拆解 LLM 交互中的隐私-效用权衡：Veilmind-4B 意图驱动本地保护框架

    提出意图驱动本地框架 Veilmind，净化发往远程模型的提示词

    arXiv
    2609.10992
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Evaluated models
    Veilmind-4B
    Abstract本地模型按意图净化提示词，在 Uprise 上给出两个隐私-效用工作点。

    本地侧的意图驱动净化，用来降低用户提示词发往不受信任远程模型时的敏感属性暴露。

    Paper analysisRead analysis
  20. 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
Load more