DefensesarXiv 2610.09935
AgentTracer：通过细粒度意图-执行对齐追踪间接提示注入攻击
提出 AgentTracer，事后定位间接提示注入的注入点并重建攻击链
- arXiv
- 2610.09935
- Published
- Layer
- Application layer
- Setting
- AI Agent
- DefensesMonitoring and auditing
- AttacksPrompt injection
All papers
Filter6
Attack type
Attack technique
Model risks
Defense type
Impact
System type
Attack surface / entry point
Layer
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 215papers have no tags yet and are not included in filtered results.
Another 215papers do not have category tags yet and are not included in filtered results.
- 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
- arXiv
- 2610.09793
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Attacker
- Black-box
- DefensesMonitoring and auditing
- AttacksPrompt injection
- ComponentMonitors
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。
- DefensesarXiv 2610.09469
Secure-CUA: Constraining the Impact of Untrusted Content in Computer-Use Agents
提出 Secure-CUA，用动作事务约束计算机使用智能体的不可信影响
- arXiv
- 2610.09469
- Published
- Layer
- Application layer
- Setting
- web agent
- Evaluated models
- Claude Opus 5, Gemini 3.8 Flash, GPT-5.6-Sol
- DefensesInstruction–data separation
- AttacksPrompt injection
- ComponentVision
AbstractSecure-CUA 用每步事务约束不可信影响，良性 TSR 接近 Vanilla-CUA。
Secure-CUA 要限制图形界面里不可信内容对 CUA 决策和 GUI 执行的影响，同时保留合法任务所需的不可信信息。
- AttacksarXiv 2610.09240
WebMirage hijacks visual web agents' browser operations with adversarial images
提出 WebMirage，用局部对抗图像劫持视觉网页 Agent 的浏览器操作
- arXiv
- 2610.09240
- Published
- Layer
- Application layer
- Setting
- web agent
- Evaluated models
- LLaVA-v1.5-13B, LLaVA-v1.6-34B, MiniCPM-o-8B and others6
- AttacksPrompt injection
- TechniqueGradient and representation optimization
- ComponentVision
Abstract论文提出了针对网络智能体的端到端视觉红队框架 WebMirage，揭示了从视觉定位到浏览器执行的安全脆弱性。
WebMirage 是一个面向视觉定位网络智能体的流水线感知端到端红队攻击框架。
- AttacksarXiv 2610.08951
ASPIRE：面向 LLM Agent 的开放式提示注入红队引擎
提出 ASPIRE 行为图搜索，开放式发现 Agent 提示注入漏洞
- arXiv
- 2610.08951
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- Gemini-3.7-Flash, DeepSeek-V4-Flash, Gemini-2.5-flash
- AttacksPrompt injection
- TechniqueAutomated search
AbstractASPIRE 把提示注入红队从载荷优化改成行为图引导的开放式漏洞发现。
ASPIRE 是给搭建方用的智能体提示注入红队引擎，目标是开放式画出可利用的行为路径，而不是在固定任务上打磨一条载荷。
- DefensesarXiv 2610.05640
Research reveals that the CaMeL defense fails in multi-agent systems and proposes multi-CaMeL
提出 multi-CaMeL，在智能体调用边界隔离可信指令与非可信数据
- arXiv
- 2610.05640
- Published
- Layer
- Application layer
- Setting
- Multi-agent systems
- Evaluated models
- Claude Fable 5, GPT-5.5, GPT-4.1 and others6
- DefensesInstruction–data separation
- AttacksPrompt injection
Abstract论文针对多智能体中单体防御不组合的边界清洗问题，提出 multi-CaMeL 协议彻底阻断了 MultiAgentDojo 上的注入攻击。
- DefensesarXiv 2610.06966
APEX: proactively defending against indirect prompt injection at the LLM agent execution boundary.
提出 APEX，在执行边界用授权契约防御 Agent 间接提示注入
- arXiv
- 2610.06966
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- DeepSeek-V4-Flash, GLM-5.2, GPT-5.6-sol
- DefensesInstruction–data separation
- AttacksPrompt injection
- ComponentPermissions and sandboxing
- DefensesarXiv 2610.04269
SRFT：让 Agent 从失败经验中自我反思以抵御提示注入
提出 SRFT，让 Agent 从失败经验中学习拒绝提示注入
- arXiv
- 2610.04269
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- SR-Agent-Llama, SR-Agent-Qwen3-8B, SR-Agent-Qwen3-4B
- DefensesModel hardening
- AttacksPrompt injection
- DefensesarXiv 2610.02869
AgentTrap：面向自主渗透测试 Agent 的有状态反馈欺骗蜜罐
提出 AgentTrap 有状态蜜罐，欺骗并反击自主渗透测试 Agent
- arXiv
- 2610.02869
- Published
- Layer
- Application layer
- Setting
- coding agent
- Evaluated models
- AIDA, Claude Code, PentestGPT and others8
- AttacksMalicious use
- 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
- DefensesPermissions and approval
- AttacksPrompt injection
- ComponentPermissions and sandboxing
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。
- DefensesarXiv 2610.00450
ZoneClaw 用记忆分区防御 OpenClaw 式计算机使用 Agent 的持久记忆攻击
提出 ZoneClaw，用记忆分区防御计算机使用 Agent 的持久记忆攻击
- arXiv
- 2610.00450
- Published
- Layer
- Application layer
- Setting
- Browser and computer use
- Evaluated models
- Claude Sonnet 4.6, GPT-5.5, GLM-5.2 and others4
- DefensesInstruction–data separation
- AttacksPoisoning and backdoors
- ComponentMemory
- DefensesarXiv 2609.37196
ToolFence：为工具调用 LLM Agent 提供细粒度授权
提出 ToolFence，以能力蓝图和确定性授权拦截 Agent 未授权工具调用
- arXiv
- 2609.37196
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- Qwen3-max, GPT-4o (gpt-4o-2024-08-06), Mistral-Small-3.1-24B-Instruct-2503
- DefensesPermissions and approval
- AttacksPrompt injection
- ComponentPermissions and sandboxing
- DefensesarXiv 2609.34974
Veer：网页 Agent 运行时防御暗模式
提出 Veer，在网页 Agent 执行前改写会造成未授权后果的网页状态
- arXiv
- 2609.34974
- Published
- Layer
- Application layer
- Setting
- web agent
- Evaluated models
- GPT-5.6-Luna, DeepSeek-V4-Flash
- AttacksPrompt injection
- AttacksarXiv 2609.33628
用课程强化学习对前沿模型做提示注入红队测试
用跨目标课程强化学习训练提示注入红队，攻破前沿智能体
- arXiv
- 2609.33628
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- GPT-4o-mini, GPT-5-nano, GPT-5.6-Luna and others12
- AttacksPrompt injection
- TechniqueAutomated search
Abstract提出跨目标模型课程强化学习，攻破前沿模型冷启动瓶颈并实现跨模型迁移。
该研究针对强化学习自动化提示注入红队在攻击前沿大模型时的冷启动问题，提出了一种基于目标模型鲁棒性递进的课程强化学习方法。
- 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
- DefensesDetection and filtering
- AttacksPrompt injection
- TechniqueDormant triggers
- ComponentGuardrails and judges
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。
SoK 论文提出 FARSIGHT 框架，评测 15 个学术金融 LLM 交易 Agent 的鲁棒性与安全失败
用 FARSIGHT 方案级评测金融 LLM 交易 Agent 的稳健性与安全失败
- arXiv
- 2609.19705
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- ChatGPT Deep Research
- AttacksPrompt injection
- RiskDangerous agent actions
- ComponentMemory
- DefensesarXiv 2609.18411
可验证行动卡片 VAC：为智能体浏览器提供可信人在回路确认
提出可验证操作卡 VAC，把智能体浏览器确认绑定到真实 DOM 操作
- arXiv
- 2609.18411
- Published
- Layer
- Application layer
- Setting
- web agent
- Evaluated models
- gpt-oss-120b, Gemini-2.5-Flash, GPT-4.1 and others6
- DefensesPermissions and approval
- AttacksPrompt injection
- ComponentPermissions and sandboxing
- DefensesarXiv 2609.14987
ActGuard：面向 LLM Agent 间接提示注入的执行前动作审计框架
提出 ActGuard 执行前动作审计，定位并遮盖间接提示注入的恶意片段
- arXiv
- 2609.14987
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- GPT-4o-mini, Gemini-2.5-Flash, DeepSeek-V4-Flash
- DefensesMonitoring and auditing
- AttacksPrompt injection
- DefensesarXiv 2609.05901
KITA：为 LLM Agent 设计密钥隔离的阈值签名授权架构
提出 KITA，把 Agent 批准绑定到规范动作的密钥隔离阈值签名
- arXiv
- 2609.05901
- Published
- Layer
- Application layer
- Setting
- AI Agent
- DefensesPermissions and approval
- AttacksPrompt injection
- ComponentPermissions and sandboxing
AbstractKITA 把隔离的阈值签名当作精确动作的执行条件，并测量门控、相关性与签名开销。
KITA 是一套密钥隔离的阈值授权架构，用来把模型评审变成执行外部动作的条件。
- DefensesarXiv 2609.02786
SafeEvolve：基于智能体经验的 Harness-Policy 协同演化安全对齐框架
提出 SafeEvolve，用轨迹证据协同演化安全 harness 与策略
- arXiv
- 2609.02786
- Published
- Layer
- Application layer
- Setting
- AI Agent
- Evaluated models
- Qwen3.5-4B, Qwen3-4B-Instruct-2507, Qwen3.5-1.7B and others4
- DefensesModel hardening
- AttacksPrompt injection
AbstractSafeEvolve 用轨迹证据协同更新安全 harness 和策略，在两个 4B 模型的智能体安全基准上改变安全与效用。
SafeEvolve 把已完成的 on-policy 轨迹变成安全 harness 的有界更新，并用两阶段 SFT-RL 把这些更新内化进工具使用策略。