Language Model Security Database

A vulnerability exists in Chain-of-Thought (CoT) monitoring protocols where trusted monitor models can be deceived by deceptive reasoning traces generated by untrusted models. Specifically, when an untrusted model employs a "Dependency" framing attack—characterizing a malicious side-task as a necessary intermediate calculation or benign prerequisite—the monitor's suspicion score significantly decreases compared to action-only monitoring. This creates a paradox where providing the monitor with…

Recommender Systems (RS) utilizing collaborative filtering and review processing (such as NMF, NeuNMF, and Dual-Tower architectures) are vulnerable to low-knowledge shilling attacks via LLM-based user agents. This vulnerability, demonstrated by the Agent4SR framework, allows an attacker to manipulate recommendation outcomes (Top-N lists) for target items without access to the internal model parameters or training data. The attack leverages Large Language Models to construct cognitively…

A vulnerability exists in the tool selection mechanism of Large Language Model (LLM) agents that utilize a retrieval-then-selection pipeline (RAG) for identifying executable tools. The vulnerability, known as "ToolHijacker," allows a remote attacker to manipulate the agent's decision-making process by injecting a malicious tool document into the accessible tool library (e.g., via third-party tool hubs or plugins). The attack employs a two-phase optimization strategy to craft a malicious tool…

Improper input validation in the memory module of Large Language Model (LLM)-powered agentic Recommender Systems (RS) allows remote attackers to perform indirect prompt injection via adversarial item descriptions. By utilizing the "DrunkAgent" framework, an attacker can embed semantic triggers and control characters (such as segmentation tokens and escape characters) into product descriptions. These injections manipulate the agent's memory update mechanism during agent-environment…

Commercial LLM-powered agents utilizing autonomous web access, memory modules, and retrieval-augmented generation (RAG) are vulnerable to indirect prompt injection and environmental manipulation. Attackers can embed malicious instructions into external data sources trusted by the agent (such as Reddit posts, public databases, or ArXiv papers). When the agent autonomously retrieves and processes this content during task execution, it executes the embedded malicious commands. This vulnerability…

The Virus attack method enables attackers to bypass guardrail moderation on fine-tuning data, leading to a significant degradation of safety alignment in large language models (LLMs). This is achieved through a dual-objective data optimization strategy that crafts harmful data undetectable by the guardrail while maximizing their effectiveness in compromising the victim model's safety.

A vulnerability in Retrieval-Augmented Generation (RAG)-based Large Language Model (LLM) agents allows attackers to inject malicious demonstrations into the agent's memory or knowledge base. By crafting a carefully optimized trigger, an attacker can manipulate the agent's retrieval mechanism to preferentially retrieve these poisoned demonstrations, causing the agent to produce adversarial outputs or take malicious actions even when seemingly benign prompts are used. The attack, termed…