Skip to main content
LLM Security Database
Skip to research search
Last analyzed Sep 9, 2026

Language Model Security Database

985 research findings · 1119 evaluated models

Filtered research findings

798 entries

Matches every word across titles, descriptions, sources, affected systems, and models.

Embodied AI Policy Jailbreak

Published Analyzed Source: arXiv

LLM-based planning modules in embodied AI systems are vulnerable to Policy Executable (POEX) jailbreak attacks. Attackers can inject carefully crafted adversarial suffixes into user instructions, causing the LLM to generate and execute harmful policies in both simulated and real-world environments. The attacks bypass safety mechanisms by using optimized, human-readable suffixes that evade perplexity-based detection.

POEX: Policy Executable Embodied AI Jailbreak Attacks
Evaluated models: Claude 3.5 Sonnet, GPT-4, GPT-4 Turbo+9 more
  • Claude 3.5 Sonnet
  • GPT-4
  • GPT-4 Turbo
  • GPT-4o
  • GPT-4o Mini
  • Llama 3 8B
  • Llama 3 8B Instruct
  • Llama 3.1 70B Instruct
  • Mistral 7B Instruct v0.2
  • Mixtral 8x22B Instruct
  • Phi-3
  • Vicuna 13B v1.5

Multimodal LLM Jailbreak

Published Analyzed Source: arXiv

A hybrid multimodal jailbreaking attack, dubbed JMLLM, exploits vulnerabilities in 13 popular large language models (LLMs) across text, image, and speech modalities. The attack leverages alternating translation, word encryption, feature collapse in images, and harmful text injection to bypass safety mechanisms and elicit harmful responses. Success rates vary across LLMs and modalities, with some models exhibiting significantly higher vulnerability than others.

Divide and Conquer: A Hybrid Strategy Defeats Multimodal Large Language Models
Evaluated models: Claude 1, Claude 2, ERNIE 3.5 Turbo+10 more
  • Claude 1
  • Claude 2
  • ERNIE 3.5 Turbo
  • GPT-3.5 Turbo
  • GPT-4
  • GPT-4o
  • GPT-4o Mini
  • Llama 2 7B
  • Llama 3 8B
  • Llama 3 70B
  • Llama 3.1 405B
  • Qwen 2.5 72B
  • Qwen VL Max

Contextual Adversarial Prompts

Published Analyzed Source: arXiv

Large Language Models (LLMs) are vulnerable to human-readable adversarial prompts crafted using situational context derived from movie scripts. These prompts, which combine a malicious prompt, a seemingly innocuous adversarial insertion, and relevant contextual information, can bypass LLMs' safety mechanisms and elicit harmful responses. The technique leverages the LLM's ability to understand context and generate responses consistent with that context to mask the malicious intent. The…

Human-Readable Adversarial Prompts: An Investigation into LLM Vulnerabilities Using Situational Context
Evaluated models: Flan-t5 Large, Gemini 1.5 Pro, Gemma 2B IT+16 more
  • Flan-t5 Large
  • Gemini 1.5 Pro
  • Gemma 2B IT
  • Gemma 7B
  • Gemma 7B IT
  • GPT-3.5 Turbo
  • GPT-4
  • GPT-4 Turbo
  • Llama 2 13B Chat
  • Llama 2 7B
  • Llama 3.1 8B
  • Llama 2 7B Chat
  • Meta-llama-3-8B
  • Mistral 7B Instruct v0.2
  • Mistral-7B-v0.1
  • Mistral-8x7B-instruct-v0.1
  • Phi-1.5
  • Phi-3-mini-128k-instruct
  • Vicuna 13B v1.5

Preference-Optimized Jailbreak

Published Analyzed Source: arXiv

JailPO is a black-box attack framework that leverages preference optimization to generate effective jailbreak prompts for aligned LLMs. The attack automatically generates prompts, bypassing safety mechanisms and eliciting harmful or undesirable responses from the target LLM. The framework includes three attack patterns (QEPrompt, TemplatePrompt, MixAsking) with varying degrees of effectiveness and risk.

JailPO: A Novel Black-box Jailbreak Framework via Preference Optimization against Aligned LLMs
Evaluated models: GPT-3.5 Turbo

Linked-Task LLM Jailbreak

Published Analyzed Source: arXiv

A novel jailbreak paradigm, Simple Assistive Task Linkage (SATA), circumvents LLM safeguards by masking harmful keywords in a malicious query and using a secondary, simple assistive task (e.g., masked language modeling or element lookup by position) to convey the masked keywords' semantics to the LLM. This distracts the LLM and allows it to bypass safety checks, leading to the generation of harmful responses.

SATA: A Paradigm for LLM Jailbreak via Simple Assistive Task Linkage
Evaluated models: Claude-v2, GPT-3.5 Turbo, GPT-4o+3 more
  • Claude-v2
  • GPT-3.5 Turbo
  • GPT-4o
  • GPT-4o Mini
  • Llama 3 70B
  • Llama 3 8B

One-Step Model Jailbreak

Published Analyzed Source: arXiv

A vulnerability in LLMs allows attackers to bypass safety mechanisms by crafting prompts that disguise malicious intent as a "defense" against harmful content. The attack, Reverse Embedded Defense Attack (REDA), leverages the model's own defensive capabilities to generate harmful outputs while masking the malicious intent within the response structure. This allows for successful jailbreaks in a single iteration, without requiring model-specific prompt engineering.

Jailbreaking? One Step Is Enough!
Evaluated models: GLM 4 9B Chat, GPT-3.5, Llama 2 13B+3 more
  • GLM 4 9B Chat
  • GPT-3.5
  • Llama 2 13B
  • Llama 3.1 8B Instruct
  • Qwen 2 7B Instruct
  • Vicuna 13B v1.5

Agent Action Hijacking

Published Analyzed Source: arXiv

A vulnerability in LLM-based agents, dubbed AI Agent Injection (AI²), allows attackers to hijack the agent's actions by manipulating the agent's memory retrieval mechanism. The attack involves two main steps: (1) Stealing action-aware knowledge from the agent's memory using crafted adversarial queries targeting the retriever module and (2) Generating Trojan prompts consisting of a Trojan string and hijacking instructions. The Trojan string is designed to manipulate the retriever into…

Towards Action Hijacking of Large Language Model-based Agent
Evaluated models: BERT, GPT-3, GPT-4+2 more
  • BERT
  • GPT-3
  • GPT-4
  • M3E
  • MiniLM

AdvPrefix Jailbreak Enhancement

Published Analyzed Source: arXiv

Large Language Models (LLMs) employing prefix-forcing safety measures are vulnerable to jailbreak attacks if the set of "safe" prefixes is insufficiently diverse or does not account for model-specific response styles. Attackers can leverage this by crafting prompts that elicit alternative prefixes, effectively bypassing the intended safety mechanisms. The vulnerability stems from over-reliance on a limited set of prefixes (e.g., "Sure, here is...") and a failure to generalize safety mechanisms…

AdvPrefix: An Objective for Nuanced LLM Jailbreaks
Evaluated models: Not reported

Audio Adversarial Jailbreak

Published Analyzed Source: arXiv
Research ID e50b7819

Large Audio-Language Models (LALMs) are vulnerable to a stealthy adversarial jailbreak attack, AdvWave, which leverages a dual-phase optimization to overcome gradient shattering caused by audio discretization. The attack crafts adversarial audio by adding perceptually realistic environmental noise, making it difficult to detect. The attack also dynamically adapts the adversarial target based on the LALM's response patterns.

AdvWave: Stealthy Adversarial Jailbreak Attack against Large Audio-Language Models
Evaluated models: GPT-4o, Llama Omni, Qwen 2 Audio+1 more
  • GPT-4o
  • Llama Omni
  • Qwen 2 Audio
  • SpeechGPT

Index-Gradient LLM Jailbreak

Published Analyzed Source: arXiv

Large Language Models (LLMs) are vulnerable to optimization-based jailbreaking attacks that exploit gradients during the iterative process of generating adversarial suffixes. The vulnerability stems from the inefficient exploration of the token space in existing methods like Greedy Coordinate Gradient (GCG), which uniformly samples tokens for replacement regardless of gradient values. This leads to redundant computations and a slow optimization process.

Exploiting the Index Gradients for Optimization-Based Jailbreaking on Large Language Models
Evaluated models: Vicuna 7B, Guanaco 7B, Llama 2 7B Chat+5 more
  • Vicuna 7B
  • Guanaco 7B
  • Llama 2 7B Chat
  • Mistral 7B Instruct v0.2
  • GPT-3.5 Turbo
  • GPT-4
  • GPT-4o
  • Claude 3 Haiku

Research methodology

Entries summarize publicly available primary-source security research. Model names reflect only systems explicitly evaluated by the cited paper, and measurements are research-reported unless independent verification is stated.

Discovery and publication are separate: new papers enter a candidate queue, then require human review and explicit approval before publication. Updates follow that review process, not a guaranteed daily publishing schedule. “Last analyzed” is the newest analysis date among published entries; it does not mean that all recent research has been covered.