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

Oct 13 · Tuesday

All papers

Found 70 papers

Filter3

Attack type

Jailbreaks8 Prompt injection57 Poisoning and backdoors23 Detection evasion3

Attack technique

Gradient and representation optimization6 Automated search24 Encoding and obfuscation8 Role-play and persuasion3 Conversation structure manipulation3 Reasoning-chain manipulation4 Cross-modal payloads4 Defense-targeted7 Dormant triggers23

Defense type

Detection and filtering3 Instruction–data separation4

Impact

Harmful content6 Unauthorized actions27 Data and privacy leaks11 Output manipulation24

System component

MCP and skills9 Memory23 RAG5 Guardrails and judges6 Reasoning chains5 Permissions and sandboxing4 Vision4 Multimodal4

Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 201papers have no tags yet and are not included in filtered results.

Attack surface / entry point：Web pages and interfaces Attack surface / entry point：Tool responses Attack surface / entry point：Memory writes Clear all

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

AttacksarXiv 2610.12292Oct 9
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
Oct 9
Layer
Application layer
Setting
LLM applications
Attacker
Black-box
Evaluated models
LAYA-TD, LAYA-EN, LAYA-ML and others5
- AttacksDetection evasion
- ComponentGuardrails and judges
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
AttacksarXiv 2610.11932Oct 8
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
Oct 8
Layer
Application layer
Setting
LLM applications
Evaluated models
Grok, Doubao, ChatGPT and others10
- AttacksPoisoning and backdoors
- ComponentRAG
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
AttacksarXiv 2610.12524Oct 8
An exploration of attacks on AI agent systems
Retest and extend prompt injection against AI agents, and use Agent Card deception to manipulate A2A selection
arXiv
2610.12524
Published
Oct 8
Layer
Application layer
Setting
AI Agent
Evaluated models
Ollama Mistral 0.9.5, llama3-8b-8192, gemini-2.0-flash-lite and others5
- AttacksPrompt injection
- TechniqueEncoding and obfuscation
- ComponentMCP and skills
AbstractSingle-agent filtering and classification can be bypassed by format or wording, and A2A card selection is also biased toward names and descriptions.
The authors retested prompt injection on their own LangChain, AutoGen, and A2A systems, and modified the Agent Card to see if choices shifted. The abstract states the experiments mainly highlight single-agent weaknesses while extending attacks to multi-agent.
Paper analysisRead analysis
AttacksarXiv 2610.09973Oct 7
从期望危害到似然：LLM 智能体越狱的概率重构
提出 OPUR 攻击，用似然梯度优化对抗后缀越狱 LLM 智能体
arXiv
2610.09973
Published
Oct 7
Layer
Application layer
Setting
AI Agent
Attacker
White-box
Evaluated models
Llama-3.1-8B-Instruct, Ministral-8B-Instruct-2410
- AttacksJailbreaks
- TechniqueGradient and representation optimization
Paper analysisRead analysis
AttacksarXiv 2610.09240Oct 7
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
Oct 7
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。
Paper analysisRead analysis
AttacksarXiv 2610.08951Oct 7
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
Oct 7
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。
Paper analysisRead analysis
AttacksarXiv 2610.07645Oct 6
SkillPoison：用成功经验投毒自进化 Agent 的技能形成
提出 SkillPoison，用局部有效的成功经验诱导自进化 Agent 形成过度泛化技能
arXiv
2610.07645
Published
Oct 6
Layer
Application layer
Setting
AI Agent
Evaluated models
deepseek-v4-flash, Qwen V3.8 Flash, GPT-5.5
- AttacksPoisoning and backdoors
- TechniqueDormant triggers
- ComponentMemory
AbstractSkillPoison 通过局部有效且验证成功的经验诱导技能过度泛化，在多个评测基准上实现了持久的下游攻击影响。
SkillPoison 提出了一种针对自进化智能体经验到技能形成过程的攻击范式，通过组织已验证的成功经验诱导技能过度泛化。
Paper analysisRead analysis
AttacksarXiv 2610.04589Oct 3
StegoMemory red team testing: agent persistent memory can be used as a cross-session covert channel
提出 StegoMemory，评测智能体持久记忆作为跨会话隐写信道
arXiv
2610.04589
Published
Oct 3
Layer
Application layer
Setting
AI Agent
Evaluated models
Llama 3.1 405B, Claude 3.5 Sonnet, GPT-4o and others13
- AttacksPoisoning and backdoors
- TechniqueEncoding and obfuscation
- ComponentMemory
Abstract研究揭示智能体持久记忆可作为隐蔽信道跨会话传输机密，写入持久化是主要防御瓶颈，呼吁建立记忆完整性防御。
Paper analysisRead analysis
AttacksarXiv 2610.04195Oct 3
MemLeak research: multi-tenant agents sharing vector memory stores have cross-user memory leakage
提出 MemLeak，评估多租户 Agent 共享向量记忆的跨用户泄露
arXiv
2610.04195
Published
Oct 3
Layer
Application layer
Setting
AI Agent
Evaluated models
Gemini 2.5 Flash, Claude Sonnet 4.5, all-MiniLM-L6-v2 and others5
- AttacksPoisoning and backdoors
- TechniqueDormant triggers
- ComponentMemory
Paper analysisRead analysis
AttacksarXiv 2609.36576Sep 29
AdaLCPI: Making agents reconstruct indirect prompt injection from long-context fragments
提出 AdaLCPI，把恶意指令拆成碎片嵌入工具返回并诱导 Agent 越权执行
arXiv
2609.36576
Published
Sep 29
Layer
Application layer
Setting
AI Agent
Evaluated models
GPT-4.1, GPT-5.1, GPT-5.6-Luna and others7
- AttacksPrompt injection
- TechniqueEncoding and obfuscation
Paper analysisRead analysis
AttacksarXiv 2609.35576Sep 29
研究提出跨 LLM 助手的记忆跳跃攻击
提出跨独立 Agent 的记忆跳跃攻击，经共享工件在助手间自传播
arXiv
2609.35576
Published
Sep 29
Layer
Application layer
Setting
AI Agent
Evaluated models
GPT-5.6 Luna, Kimi-K2.6, GPT-OSS-120B and others6
- AttacksPoisoning and backdoors
- TechniqueDormant triggers
- ComponentMemory
Abstract论文证实持久工件与内存可形成跨智能体传播链，揭示了将共享工件纳入安全防线的重要意义。
该论文系统研究并量化了通过共享持久工件在具有独立私有内存的 LLM 智能体之间发生自传播的“工件介导传播”攻击风险。
Paper analysisRead analysis
AttacksarXiv 2609.33628Sep 27
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
Sep 27
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。
Paper analysisRead analysis
AttacksarXiv 2609.31318Sep 25
AgentXploit：面向 AI Agent 的仓库到运行时自动化红队系统
提出 AgentXploit，从代码仓库发现攻击路径并在运行时验证利用
arXiv
2609.31318
Published
Sep 25
Layer
Application layer
Setting
AI Agent
Evaluated models
GPT-4.1
- AttacksPrompt injection
Paper analysisRead analysis
AttacksarXiv 2609.28613Sep 24
研究：提示注入可改变 Jev 决策模型的行动概率
测量间接提示注入对类型化概率决策的动作劫持
arXiv
2609.28613
Published
Sep 24
Layer
Application layer
Setting
Models and APIs
Evaluated models
jev-1.13.0
- AttacksPrompt injection
- TechniqueAutomated search
Abstract封闭动作集改变提示注入的表现，但没有去掉概率被不可信内容移动的风险。
作者在 jev-1.13.0 上研究 decision hijacking：不可信内容能否在用户任务和可选动作都不变时，把类型化决策推向攻击者目标。
Paper analysisRead analysis
AttacksarXiv 2609.28585Sep 23
论文提出持久计费状态概念，揭示工具调用 LLM Agent 的拒绝钱包攻击与防御
定义工具调用 Agent 的持久计费状态，提出拒绝钱包攻击与防御
arXiv
2609.28585
Published
Sep 23
Layer
Application layer
Setting
AI Agent
Evaluated models
GPT-4o, GPT-4o-mini, Mistral-Large and others7
- DefensesInstruction–data separation
- AttacksDenial of service
Abstract已准入工具返回的跨轮保留会重复计费。
这篇工作把工具调用智能体里跨轮保留、并被提供商反复计费的外部工具返回定义为 persistent billable state，并把它的主机侧再次送入决定当作控制点。。
Paper analysisRead analysis
AttacksarXiv 2609.27542Sep 23
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
arXiv
2609.27542
Published
Sep 23
Layer
Application layer
Setting
AI Agent
Evaluated models
openai/gpt-oss-20b, google/gemma-4-26B-A4B-it, deepseek-ai/DeepSeek-R1-Distill-Qwen-7B and others4
- AttacksPrompt injection
- TechniqueConversation structure manipulation
- ComponentReasoning chains
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
AttacksarXiv 2609.26761Sep 23
A2M: A two-stage agent hijacking framework targeting the MCP ecosystem
提出两阶段框架 A2M，优化 MCP 工具元数据与返回载荷以劫持 Agent
arXiv
2609.26761
Published
Sep 23
Layer
Application layer
Setting
AI Agent
Evaluated models
GLM-4.6, Qwen3-Max, DeepSeek-V3.1 and others5
- AttacksPrompt injection
- TechniqueAutomated search
- ComponentMCP and skills
Paper analysisRead analysis
AttacksarXiv 2609.22510Sep 19
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
Sep 19
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。
Paper analysisRead analysis
AttacksarXiv 2609.18217Sep 16
研究：跨通道碎片化攻击可绕过 MCP 单通道防御
提出跨通道分段攻击，利用 MCP 隐式信任外发敏感文件
arXiv
2609.18217
Published
Sep 16
Layer
Application layer
Setting
coding agent
Evaluated models
GPT-4o, GPT-4o-mini, GPT-5.4 and others15
- AttacksPrompt injection
- TechniqueEncoding and obfuscation
- ComponentMCP and skills
Abstract论文揭示了 MCP 多通道隐式信任层级，通过跨通道分段与协议级利用突破现有防御，展示了严重的凭证外发风险。
Paper analysisRead analysis
AttacksarXiv 2609.11757Sep 11
Researchers disclose a Whisper prompt injection attack on the AP2 payment protocol and propose the A-VIP defense
提出 AP2 支付协议的 Whisper 攻击与 A-VIP 绑定防御
arXiv
2609.11757
Published
Sep 11
Layer
Application layer
Setting
AI Agent
Evaluated models
gemini-2.5-flash-lite, gemini-3.1-flash-lite, gemini-3-flash-preview and others15
- DefensesPermissions and approval
- AttacksPrompt injection
- ComponentPermissions and sandboxing
Abstract论文揭示了 AP2 协议决策层受操纵的风险，提出结构化与权能绑定防御 A-VIP，并发布评测基准。
Paper analysisRead analysis

AI Safety Daily

Read the latest issueOne issue per day, updated as events develop

All papers

Attack type

Attack technique

Defense type

Impact

System type

Attacker access

Attack surface / entry point

Layer

System component

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

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

An exploration of attacks on AI agent systems

从期望危害到似然：LLM 智能体越狱的概率重构

WebMirage hijacks visual web agents' browser operations with adversarial images

ASPIRE：面向 LLM Agent 的开放式提示注入红队引擎

SkillPoison：用成功经验投毒自进化 Agent 的技能形成

StegoMemory red team testing: agent persistent memory can be used as a cross-session covert channel

MemLeak research: multi-tenant agents sharing vector memory stores have cross-user memory leakage

AdaLCPI: Making agents reconstruct indirect prompt injection from long-context fragments

研究提出跨 LLM 助手的记忆跳跃攻击

用课程强化学习对前沿模型做提示注入红队测试

AgentXploit：面向 AI Agent 的仓库到运行时自动化红队系统

研究：提示注入可改变 Jev 决策模型的行动概率

论文提出持久计费状态概念，揭示工具调用 LLM Agent 的拒绝钱包攻击与防御

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

A2M: A two-stage agent hijacking framework targeting the MCP ecosystem

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

研究：跨通道碎片化攻击可绕过 MCP 单通道防御

Researchers disclose a Whisper prompt injection attack on the AP2 payment protocol and propose the A-VIP defense