The paper introduces Persona-Conditioned Adversarial Prompting (PCAP), a method that significantly improves LLM red-teaming by simulating diverse attacker personas, leading to the discovery of more co…

The paper introduces RoleCDE, a novel benchmark that evaluates role-playing agents' ability to resolve conflicts between role-specific values and general alignment constraints, revealing a 'Role Value…

This paper introduces a novel attack, RA-ICA, that targets RAG-enhanced LLMs by poisoning external knowledge bases to drastically increase inference costs, achieving up to a 13.12x increase in token c…

AutoMIA introduces an agentic framework that automates the process of Membership Inference Attacks (MIAs) by self-exploring the attack space, achieving state-of-the-art performance without manual feat…

This paper systematically studies memory poisoning attacks in LLM agents, identifying multiple vulnerabilities and proposing a new benchmark to assess the risk.

The paper introduces Latent Reward Steering (LRS), an adaptive inference-time framework that implicitly improves the reasoning ability of LLMs by guiding the model's internal latent states based on a…

The paper introduces Persona-Conditioned Adversarial Prompting (PCAP), a novel framework that significantly enhances the discovery of jailbreaks by conditioning adversarial search on multiple attacker…

The paper introduces AgentSecBench, a security evaluation framework that measures prompt injection, privacy leakage, and tool-use integrity in LLM agents by defining formal security games and testing…

The paper introduces PrivacySIM, an evaluation suite that benchmarks how well LLMs can simulate individual user privacy decisions based on persona attributes, finding that while conditioning improves…

The paper introduces CyBiasBench, a comprehensive benchmark that quantifies the inherent, agent-specific bias in LLM agents' attack selection patterns in cybersecurity scenarios.

The paper introduces Mindgames, a comprehensive multi-game arena for evaluating LLM agents' sustained social and strategic reasoning, demonstrating that current evaluations are limited by structural s…

The paper systematically maps LLM agent vulnerabilities by testing 10,000 prompt variations, finding that 'goal reframing' language is the primary trigger for exploitation, rather than broad adversari…

This paper presents the first systematic study of black-box skill stealing attacks against proprietary LLM agents, demonstrating that structured agent skills can be easily extracted, posing a signific…

The paper introduces Critical-CoT, a novel two-stage fine-tuning defense framework that equips LLMs with critical thinking abilities to detect and reject malicious reasoning steps introduced by advanc…

This paper extends the Werewolf game with a Jester to evaluate theory-of-mind in large language models and exposes a layer of multi-agent reasoning.

The paper proposes the Layered Attack Surface Model (LASM), a structural taxonomy that maps security threats and defenses across the complex, multi-layered architecture of AI agents, revealing signifi…

The paper proposes a structured prompt engineering framework to enhance the integrity and reliability of Chain-of-Thought (CoT) reasoning in LLMs, demonstrating significant improvements in security-se…

The paper introduces an adaptive interview framework to gather rich persona context, demonstrating that LLMs improve decision alignment in moral dilemmas only when they selectively ground their decisi…

The paper proposes ADAM, a novel and highly effective privacy attack that systematically extracts sensitive data from LLM agent memory by adaptively querying the victim agent's memory based on data di…