评估多模态 AI 中的视觉隐私风险：一种基于新型分类体系的视觉-语言模型评估
Assessing Visual Privacy Risks in Multimodal AI: A Novel Taxonomy-Grounded Evaluation of Vision-Language Models
AI总结：
该研究提出多层级 Visual Privacy Taxonomy，并据此评估多种最先进 VLM，揭示其情境隐私理解的不一致，为隐私感知多模态 AI 提供分类框架与基准。
摘要：
近年来，人工智能深刻改变了技术格局。大型语言模型（LLM）在推理、文本理解、上下文模式识别以及将语言与视觉理解相结合方面展现出令人瞩目的能力。尽管这些进展带来了显著益处，但它们也暴露出模型在理解隐私概念方面的关键局限。因此，尤其是在缺乏用于测试此类任务的支持资源的情况下，人们高度关注能否以及如何让这些模型理解并执行隐私原则。在本研究中，我们通过考察法律框架如何为这些新兴技术的能力提供依据来应对这些挑战。为此，我们提出了一个全面的多层级 Visual Privacy Taxonomy（视觉隐私分类体系），涵盖广泛的隐私问题，并被设计为可扩展且能够适应现有及未来的研究需求。此外，我们评估了若干最先进的 Vision-Language Models（VLM，视觉-语言模型）的能力，揭示了它们在理解情境隐私方面存在显著不一致。我们的工作既为未来研究贡献了基础性分类体系，也提供了关于当前模型局限性的关键基准，证明迫切需要更稳健、具备隐私意识的 AI 系统。
英文摘要：
Artificial Intelligence have profoundly transformed the technological landscape in recent years. Large Language Models (LLMs) have demonstrated impressive abilities in reasoning, text comprehension, contextual pattern recognition, and integrating language with visual understanding. While these advances offer significant benefits, they also reveal critical limitations in the models' ability to grasp the notion of privacy. There is hence substantial interest in determining if and how these models can understand and enforce privacy principles, particularly given the lack of supporting resources to test such a task. In this work, we address these challenges by examining how legal frameworks can inform the capabilities of these emerging technologies. To this end, we introduce a comprehensive, multi-level Visual Privacy Taxonomy that captures a wide range of privacy issues, designed to be scalable and adaptable to existing and future research needs. Furthermore, we evaluate the capabilities of several state-of-the-art Vision-Language Models (VLMs), revealing significant inconsistencies in their understanding of contextual privacy. Our work contributes both a foundational taxonomy for future research and a critical benchmark of current model limitations, demonstrating the urgent need for more robust, privacy-aware AI systems.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
