相关论文: ConVerse: Benchmarking Contextual Safety in Agent-…

Evaluating alignment in language models requires testing how they behave under realistic pressure, not just what they claim they would do. While alignment failures increasingly cause real-world harm, comprehensive evaluation frameworks with…

Autonomous web agents solve complex browsing tasks, yet existing benchmarks measure only whether an agent finishes a task, ignoring whether it does so safely or in a way enterprises can trust. To integrate these agents into critical…

Large language models (LLMs) and LLM-based agents are increasingly deployed as assistants in planning and decision making, yet most existing systems are implicitly optimized for a single-principal interaction paradigm, in which the model is…

With the continuous development of large language models (LLMs), transformer-based models have made groundbreaking advances in numerous natural language processing (NLP) tasks, leading to the emergence of a series of agents that use LLMs as…

Large language models (LLMs) are increasingly deployed as customer-facing agents, yet evaluating their reliability remains challenging due to stochastic, multi-turn interactions. Current evaluation protocols rely on linear Monte Carlo…

The emergence of multimodal LLM-based agents (MLAs) has transformed interaction paradigms by seamlessly integrating vision, language, action and dynamic environments, enabling unprecedented autonomous capabilities across GUI applications…

In recent years, Large-Language-Model-driven AI agents have exhibited unprecedented intelligence and adaptability. Nowadays, agents are undergoing a new round of evolution. They no longer act as an isolated island like LLMs. Instead, they…

We identify and formalize a novel security risk: Context-Fragmented Violations (CFVs) - a class of policy breaches where individual agent actions appear locally safe and reasonable, yet collectively violate organizational policies because…

The rapid deployment of Large language model (LLM) agents in critical domains like healthcare and finance necessitates robust security frameworks. To address the absence of standardized evaluation benchmarks for these agents in dynamic…

As language models are increasingly deployed as autonomous agents in high-stakes settings, ensuring that they reliably follow user-defined rules has become a critical safety concern. To this end, we study whether language models exhibit…

The rapid growth of metaverse technologies, including virtual worlds, augmented reality, and lifelogging, has accelerated their adoption across diverse domains. This rise exposes users to significant new security and privacy challenges due…

In this paper, we develop a user-centric privacy framework for quantitatively assessing the exposure of personal information in open settings. Our formalization addresses key-challenges posed by such open settings, such as the unstructured…

Multi-agent systems, when enhanced with Large Language Models (LLMs), exhibit profound capabilities in collective intelligence. However, the potential misuse of this intelligence for malicious purposes presents significant risks. To date,…

Given the progressive nature of the world today, fairness is a very important social aspect in various areas, and it has long been studied with the advent of technology. To the best of our knowledge, methods of quantifying fairness errors…

The integration of vision-language models (VLMs) is driving a new generation of embodied agents capable of operating in human-centered environments. However, as deployment expands, these systems face growing safety risks, particularly when…

The metaverse promises unprecedented immersive digital experiences but also raises critical privacy concerns as vast amounts of personal and behavioral data are collected. As immersive technologies blur the boundaries between physical and…

Large language models (LLMs) now mediate many web-based mental-health, crisis, and other emotionally sensitive services, yet their psychosocial safety in these settings remains poorly understood and weakly evaluated. We present DialogGuard,…

Evolving AI systems increasingly deploy multi-agent architectures where autonomous agents collaborate, share information, and delegate tasks through developing protocols. This connectivity, while powerful, introduces novel security risks.…

Large language models (LLMs) are evolving into agentic systems that reason, plan, and operate external tools. The Model Context Protocol (MCP) is a key enabler of this transition, offering a standardized interface for connecting LLMs with…

Voice privacy approaches that preserve the anonymity of speakers modify speech in an attempt to break the link with the true identity of the speaker. Current benchmarks measure speaker protection based on signal-to-signal comparisons. In…