Skip to content

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

Oct 12 · Monday

All papers

Found 6 papers
Filter4

Attack type

Prompt injection4

Attack technique

Gradient and representation optimization3Cross-modal payloads3

Impact

Harmful content3Unauthorized actions13Data and privacy leaks4Output manipulation6Availability and reliability3

System type

LLM applications3AI Agent3

Attack surface / entry point

Web pages and interfaces3Images, audio and video3

Layer

Prompt layer27Model layer17Application layer6Training and data layer10

System component

MCP and skills5Memory22RAG27Guardrails and judges20Monitors3Reasoning chains4Permissions and sandboxing8Vision5Audio1Embeddings5Multimodal7
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.
Impact：Output manipulationLayer：Application layerSystem component：VisionSystem component：AudioClear all

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

  3. 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
  4. AttacksarXiv 2606.12716

    PaperGuard：针对多模态 AI 同行评审的攻击与防御基准

    提出 PaperGuard，用图文注入与梯度扰动抬高自动审稿分数并检测恶意指令

    arXiv
    2606.12716
    Published
    Layer
    Application layer
    Setting
    LLM applications
    Evaluated models
    Claude-sonnet-4.5, GPT-4o, GPT-4.1 and others11
    AbstractPaperGuard 测到提示注入与白盒视觉抬分，并比较分块检索防御。

    PaperGuard同时评测并检测多模态 AI 审稿中的评分操纵。

    Paper analysisRead analysis
  6. AttacksarXiv 2603.16960

    Adversarial attacks against modern vision-language models

    Conduct adversarial attacks on product images for open-source visual shopping agents and induce wrong purchases

    arXiv
    2603.16960
    Published
    Layer
    Application layer
    Setting
    web agent
    Evaluated models
    LLaVA-v1.5-7B, Qwen2.5-VL-7B
    AbstractThe authors state that LLaVA is easily induced by three attacks to buy the wrong item, while Qwen's correct purchase rate remains close to its clean condition.

    The authors compare the robustness of two open-source VLM shopping agents against image adversarial perturbations in a self-built e-commerce environment.

    Paper analysisRead analysis
You have reached the end