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

Oct 12 · Monday

All papers

Found 11 papers

Filter3

Attack type

Jailbreaks8

Attack technique

Multi-turn escalation4 Encoding and obfuscation11 Role-play and persuasion5 Cross-modal payloads4

Impact

Harmful content8

System type

Models and APIs9 AI Agent3 coding agent2

Attack surface / entry point

User input9 Images, audio and video3

Layer

Prompt layer9

System component

Multimodal4

Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 261papers have no tags yet and are not included in filtered results.

Attack technique：Encoding and obfuscation System type：Models and APIs System type：coding agent Clear all

Another 261papers do not have category tags yet and are not included in filtered results.

Evaluation and benchmarksarXiv 2610.11766Oct 8
研究提出意图恢复率作为 LLM 安全评测的补充指标
提出 IRIS，用意图恢复率区分遮蔽越狱下无害输出的原因
arXiv
2610.11766
Published
Oct 8
Layer
Prompt layer
Setting
Models and APIs
Evaluated models
GPT-4o, Claude Sonnet 4.5, Claude Opus 4.5 and others6
- AttacksJailbreaks
- TechniqueEncoding and obfuscation
Abstract无害输出的证据权重取决于任务是否被恢复。
IRIS 把意图恢复当作中间信号，用来解释意图遮蔽提示下为何没有出现有害协助。
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2610.09033Oct 7
研究发布 ASRD 数据集，评测五款开源模型在非规范输入下的安全表现
arXiv
2610.09033
Published
Oct 7
Layer
Prompt layer
Setting
Models and APIs
Evaluated models
Gemma 2 9B, Gemma 3 4B, Llama 3.1 8B and others6
- AttacksJailbreaks
- TechniqueEncoding and obfuscation
Abstract五模型、七族表面形式的四态评测里，低有害遵从不等于读懂后的安全处理。
ASRD 用四态标签检查五个开源模型如何处理非规范表面上的有害请求。
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2609.35912Sep 28
MMSkillRisk：多模态技能图像中的隐藏指令可诱导 Agent 执行未授权操作
提出 MMSkillRisk 基准，评测多模态技能图像中的隐藏指令诱导 Agent 越权操作
arXiv
2609.35912
Published
Sep 28
Layer
Application layer
Setting
coding agent
Evaluated models
GPT-5.6-sol, Kimi-K2.6, Kimi-K3 and others6
- AttacksPrompt injection
- TechniqueCross-modal payloads
- ComponentMCP and skills
Abstract论文提出 MMSkillRisk 基准与 NCVA 攻击，揭示多模态技能中教学图像潜藏指令可引发未授权操作并与良性任务共存。
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2609.07216Sep 7
DIVA：利用视觉锚点对视频生成模型实施多模态越狱
提出 DIVA，将有害意图拆成视觉锚点与运动文本以越狱视频生成模型
arXiv
2609.07216
Published
Sep 7
Layer
Prompt layer
Setting
Models and APIs
Attacker
Black-box
Evaluated models
Wan, CogVideoX, LTX-Video and others9
- AttacksJailbreaks
- TechniqueCross-modal payloads
- ComponentImage generation
AbstractDIVA 用视觉锚点抑制语义漂移做视频越狱，并给出多条件安全基准 TI2VSafetyBench。
DIVA 是针对带参考图的视频生成模型的免训练 black-box 越狱：用一致性机制本身堵住语义漂移这条安全逃逸路径。。
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2609.05843Sep 5
SinoGlyphBench：中文字形混淆下语言模型审核的诊断基准
提出 SinoGlyphBench，用配对字形混淆诊断中文审核稳定性
arXiv
2609.05843
Published
Sep 5
Layer
Prompt layer
Setting
Models and APIs
Evaluated models
GPT-5.4 mini, GPT-5.5, DeepSeek V4 Flash and others12
- AttacksDetection evasion
- TechniqueEncoding and obfuscation
- ComponentGuardrails and judges
AbstractSinoGlyphBench 显示中文字形混淆会同时增加有害漏报和误报，全文扰动最重。
SinoGlyphBench是一个配对诊断基准，用来看内容审核在可还原的中文字形混淆下是否还保持原决定。
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2608.06862Aug 7
SynChain：诱导计算机使用 Agent 自建攻击链并跨任务持久化
提出 SYNCHAIN，诱导计算机使用 Agent 自建跨任务持久攻击链
arXiv
2608.06862
Published
Aug 7
Layer
Application layer
Setting
Browser and computer use, coding agent
Evaluated models
Qwen3.5-9B, Llama-3.1-8B, Ministral-3-8B
- AttacksPoisoning and backdoors
- TechniqueDormant triggers
- ComponentMCP and skills
Abstract论文揭示了 CUA 自合成工件带来的内部供应链威胁，验证了潜伏载荷的跨任务传播能力与防御局限。
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2606.29602Jun 29
六款主流 LLM 在多语言与混淆提示注入攻击下的实证评估
评测多语言与编码混淆提示下模型生成钓鱼与键盘记录内容的遵从
arXiv
2606.29602
Published
Jun 29
Layer
Prompt layer
Setting
Models and APIs
Evaluated models
DeepSeek-V3.2, Gemini 2.5 Flash, GPT-5 Mini and others6
- AttacksJailbreaks
- TechniqueEncoding and obfuscation
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2606.06037Jun 4
SpeechJBB：语码转换语音下大型音频语言模型的越狱基准
提出 SpeechJBB，评测语码转换语音下音频语言模型的越狱稳健性
arXiv
2606.06037
Published
Jun 4
Layer
Prompt layer
Setting
Models and APIs
Evaluated models
Qwen2.5-Omni-7B, Qwen3-Omni-30B-A3B-Instruct, Voxtral-Small-24B and others9
- AttacksJailbreaks
- TechniqueEncoding and obfuscation
- ComponentAudio
AbstractSpeechJBB 用五语语码转换语音和伪词探针评测九个 LALM，作者将越狱归为对齐失败而非听不懂。
SpeechJBB 用合成语音测量 LALM 在五语、语码转换和局部伪词混淆下的安全对齐，并另设口语理解对照。
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2601.04740Jan 8
StealthGraph: Revealing Domain-Specific Risks in LLMs via Knowledge Graph-Guided Harmful Prompt Generation
提出 StealthGraph，合成领域隐式有害提示并评测模型
arXiv
2601.04740
Published
Jan 8
Layer
Prompt layer
Setting
Models and APIs
Evaluated models
GPT-4o-mini, Gemini 2.5 Flash, Grok 3 Mini and others12
- AttacksJailbreaks
- TechniqueEncoding and obfuscation
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2601.00588Jan 2
CSSBench: Evaluating the safety of lightweight LLMs against Chinese-specific adversarial patterns
提出 CSSBench 评测轻量模型对中文表面对抗改写的失守与过度拒绝
arXiv
2601.00588
Published
Jan 2
Layer
Prompt layer
Setting
Models and APIs
Evaluated models
Qwen3-0.6B, Qwen3-1.7B, Qwen3-8B and others10
- AttacksJailbreaks
- RiskInappropriate refusal
- TechniqueEncoding and obfuscation
AbstractCSSBench 用表面扰动和边界查询评轻量模型。
CSSBench 评测轻量 LLM 在中文特有表面改写下的安全，并同时测量过度拒绝。要处理的缺口是：英文向护栏和只测干净查询的中文基准，看不清同音、拼音、符号混写、零宽字符这些写法下的失守，也分不清较低风险来自细粒度判断还是粗粒度拒答。作者称轻量模型更需要这种评测。
Paper analysisRead analysis
Evaluation and benchmarksarXiv 2601.00213Jan 1
Overlooked large language model security vulnerabilities: malicious intelligent optimization algorithm requests and their jailbreaks
提出 MOBjailbreak，将恶意优化算法请求改写为无害表达实施越狱
arXiv
2601.00213
Published
Jan 1
Layer
Prompt layer
Setting
Models and APIs
Attacker
Black-box
Evaluated models
GPT-4o, GPT-5, OpenAI-o3 and others13
- AttacksJailbreaks
- TechniqueEncoding and obfuscation
Paper analysisRead analysis

You have reached the end

AI Safety Daily

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