Safety Vulnerabilities

A vulnerability in multi-agent Large Language Model (LLM) systems allows for a permutation-invariant adversarial prompt attack. By strategically partitioning adversarial prompts and routing them through a network topology, an attacker can bypass distributed safety mechanisms, even those with token bandwidth limitations and asynchronous message delivery. The attack optimizes prompt propagation as a maximum-flow minimum-cost problem, maximizing success while minimizing detection.

Multimodal Large Language Models (MLLMs) are vulnerable to a jailbreaking attack, dubbed PiCo, that leverages token-level typographic attacks on images embedded within code-style instructions. The attack bypasses multi-tiered defense mechanisms, including input filtering and runtime monitoring, by exploiting weaknesses in the visual modality's integration with programming contexts. Harmful intent is concealed within visually benign image fragments and code instructions, circumventing safety…

Large Language Models (LLMs) with user-controlled response prefilling features are vulnerable to a novel jailbreak attack. By manipulating the prefilled text, attackers can influence the model's subsequent token generation, bypassing safety mechanisms and eliciting harmful or unintended outputs. Two attack vectors are demonstrated: Static Prefilling (SP), using a fixed prefill string, and Optimized Prefilling (OP), iteratively optimizing the prefill string for maximum impact. The vulnerability…

LLM agents utilizing external tools are vulnerable to indirect prompt injection (IPI) attacks. Attackers can embed malicious instructions into the external data accessed by the agent, manipulating its behavior even when defenses against direct prompt injection are in place. Adaptive attacks, which modify the injected payload based on the specific defense mechanism, consistently bypass existing defenses with a success rate exceeding 50%.

Multi-agent systems (MAS) utilizing Large Language Model (LLM) orchestration are vulnerable to control-flow hijacking via indirect prompt injection, leading to Remote Code Execution (RCE). This vulnerability arises when a sub-agent (e.g., a file surfer or web surfer) processes untrusted input containing adversarial metadata, such as simulated error messages or administrative instructions. The sub-agent faithfully reproduces this adversarial content in its report to the orchestrator agent. The…

Large Language Models (LLMs) are vulnerable to multi-turn adversarial attacks that exploit incremental policy erosion. The attacker uses a breadth-first search strategy to generate multiple prompts at each turn, leveraging partial compliance from previous responses to gradually escalate the conversation towards eliciting disallowed outputs. Minor concessions accumulate, ultimately leading to complete circumvention of safety measures.

Autoregressive Large Language Models (LLMs) suffer from a dynamic discriminative degradation vulnerability during sequence generation. When processing complex or adversarial inputs, the model's internal capability to distinguish between benign and harmful token sequences—measured by the linear separability of their hidden states—progressively diminishes as generation continues. If an attacker successfully bypasses the model's initial safety compliance judgment (early generation steps), the…

Large Language Models (LLMs) deployed using "Batch Prompting" strategies—where multiple distinct user queries are concatenated and processed in a single inference pass to reduce computational costs—are vulnerable to Cross-Query Prompt Injection. When a batch contains a mixture of benign queries and a single malicious query, the instructions within the malicious query (e.g., "apply this rule to every answer") bleed over the context window. This causes the model to apply the adversary's…

Large Vision-Language Models (VLMs) are vulnerable to a cross-modal toxic continuation attack facilitated by reinforcement learning-tuned diffusion models. This vulnerability allows an attacker to bypass safety alignment and external guardrails (such as NSFW image filters) by pairing a specific text prefix with a "semantically adversarial" image. Unlike traditional gradient-based adversarial examples that rely on pixel noise, these images are semantically coherent but optimized via Denoising…

Fine-tuning Large Language Models (LLMs) on the CyberLLMInstruct dataset results in a critical degradation of safety alignment and refusal mechanisms. While the dataset comprises "pseudo-malicious" content (educational descriptions of malware, phishing, and exploits without executable payloads), the Supervised Fine-Tuning (SFT) process on this corpus causes the models to generalize this instruction-following behavior to actual malicious requests. This effectively bypasses safety guardrails…