The Dark Side of Trust: Authority Citation-Driven Jailbreak Attacks on Large Language Models

The widespread deployment of large language models (LLMs) across various domains has showcased their immense potential while exposing significant safety vulnerabilities. A major concern is ensuring that LLM-generated content aligns with human values. Existing jailbreak techniques reveal how this alignment can be compromised through specific prompts or adversarial suffixes. In this study, we introduce a new threat: LLMs' bias toward authority. While this inherent bias can improve the quality of outputs generated by LLMs, it also introduces a potential vulnerability, increasing the risk of producing harmful content. Notably, the biases in LLMs is the varying levels of trust given to different types of authoritative information in harmful queries. For example, malware development often favors trust GitHub. To better reveal the risks with LLM, we propose DarkCite, an adaptive authority citation matcher and generator designed for a black-box setting. DarkCite matches optimal citation types to specific risk types and generates authoritative citations relevant to harmful instructions, enabling more effective jailbreak attacks on aligned LLMs.Our experiments show that DarkCite achieves a higher attack success rate (e.g., LLama-2 at 76% versus 68%) than previous methods. To counter this risk, we propose an authenticity and harm verification defense strategy, raising the average defense pass rate (DPR) from 11% to 74%. More importantly, the ability to link citations to the content they encompass has become a foundational function in LLMs, amplifying the influence of LLMs' bias toward authority.

AdaPPA: Adaptive Position Pre-Fill Jailbreak Attack Approach Targeting LLMs

Jailbreak vulnerabilities in Large Language Models (LLMs) refer to methods that extract malicious content from the model by carefully crafting prompts or suffixes, which has garnered significant attention from the research community. However, traditional attack methods, which primarily focus on the semantic level, are easily detected by the model. These methods overlook the difference in the model's alignment protection capabilities at different output stages. To address this issue, we propose an adaptive position pre-fill jailbreak attack approach for executing jailbreak attacks on LLMs. Our method leverages the model's instruction-following capabilities to first output pre-filled safe content, then exploits its narrative-shifting abilities to generate harmful content. Extensive black-box experiments demonstrate our method can improve the attack success rate by 47% on the widely recognized secure model (Llama2) compared to existing approaches. Our code can be found at: https://github.com/Yummy416/AdaPPA.

Enhancing Cross-Prompt Transferability in Vision-Language Models through Contextual Injection of Target Tokens

Vision-language models (VLMs) seamlessly integrate visual and textual data to perform tasks such as image classification, caption generation, and visual question answering. However, adversarial images often struggle to deceive all prompts effectively in the context of cross-prompt migration attacks, as the probability distribution of the tokens in these images tends to favor the semantics of the original image rather than the target tokens. To address this challenge, we propose a Contextual-Injection Attack (CIA) that employs gradient-based perturbation to inject target tokens into both visual and textual contexts, thereby improving the probability distribution of the target tokens. By shifting the contextual semantics towards the target tokens instead of the original image semantics, CIA enhances the cross-prompt transferability of adversarial images.Extensive experiments on the BLIP2, InstructBLIP, and LLaVA models show that CIA outperforms existing methods in cross-prompt transferability, demonstrating its potential for more effective adversarial strategies in VLMs.

Chain of Attack: a Semantic-Driven Contextual Multi-Turn attacker for LLM

Large language models (LLMs) have achieved remarkable performance in various natural language processing tasks, especially in dialogue systems. However, LLM may also pose security and moral threats, especially in multi round conversations where large models are more easily guided by contextual content, resulting in harmful or biased responses. In this paper, we present a novel method to attack LLMs in multi-turn dialogues, called CoA (Chain of Attack). CoA is a semantic-driven contextual multi-turn attack method that adaptively adjusts the attack policy through contextual feedback and semantic relevance during multi-turn of dialogue with a large model, resulting in the model producing unreasonable or harmful content. We evaluate CoA on different LLMs and datasets, and show that it can effectively expose the vulnerabilities of LLMs, and outperform existing attack methods. Our work provides a new perspective and tool for attacking and defending LLMs, and contributes to the security and ethical assessment of dialogue systems.

Hierarchical Interaction Networks with Rethinking Mechanism for Document-level Sentiment Analysis

Document-level Sentiment Analysis (DSA) is more challenging due to vague semantic links and complicate sentiment information. Recent works have been devoted to leveraging text summarization and have achieved promising results. However, these summarization-based methods did not take full advantage of the summary including ignoring the inherent interactions between the summary and document. As a result, they limited the representation to express major points in the document, which is highly indicative of the key sentiment. In this paper, we study how to effectively generate a discriminative representation with explicit subject patterns and sentiment contexts for DSA. A Hierarchical Interaction Networks (HIN) is proposed to explore bidirectional interactions between the summary and document at multiple granularities and learn subject-oriented document representations for sentiment classification. Furthermore, we design a Sentiment-based Rethinking mechanism (SR) by refining the HIN with sentiment label information to learn a more sentiment-aware document representation. We extensively evaluate our proposed models on three public datasets. The experimental results consistently demonstrate the effectiveness of our proposed models and show that HIN-SR outperforms various state-of-the-art methods.