发表机构
University of Washington; Tsinghua University; Beijing Normal University; National University of Singapore; Institute of Software, Chinese Academy of Sciences; Institute for Network Sciences and Cyberspace, Tsinghua University（华盛顿大学; 清华大学; 北京师范大学; 新加坡国立大学; 中国科学院软件研究所; 清华大学网络科学与网络空间研究院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究通过二元研究和AI模型评估，发现人类共同所有者高估许可需求但熟悉度降低错位，而AI模型在预测数据可接受性上显著逊色，凸显了隐私感知AI的设计挑战。
AI 中文摘要
AI助手日益代表其主要用户调解人际沟通，但它们可能违反第三方信息所有者的隐私期望。解决这些紧张关系需要理解人类如何预期人际隐私边界。因此，我们进行了一项二元研究（N=76）以及对AI模型在18种信息类型和3种接收者关系下的匹配评估。我们发现，数据所有者的隐私判断高度依赖情境和关系。虽然熟悉的数据共同所有者与所有者的期望表现出有意义的一致性，但他们显著高估了获得许可的必要性。有趣的是，所有者与共同所有者二元组内更高的熟悉度与更高的披露可接受性和更低的共同所有者错位相关，而我们探索性的四项共情测量则不然。相比之下，即使提供了二元组内的示例，AI模型在预测数据可接受性方面也显著逊色于人类共同所有者。这些发现强调了人机交互设计中的一个核心挑战，即开发尊重多利益相关者信息边界的隐私感知AI。
英文摘要
AI assistants increasingly mediate interpersonal communication on behalf of their primary user, but they risk violating the privacy expectations of third-party information owners. Resolving these tensions requires understanding how humans anticipate interpersonal privacy boundaries. Therefore, we conducted a dyadic study (N=76) and a matched evaluation of AI models across 18 information types and 3 recipient relationships. We found that data owners' privacy judgments are highly contextual and relationship dependent. While familiar data co-owners show meaningful alignment with owners' expectations, they significantly overestimate the need for permission. Interestingly, greater familiarity within the owner-co-owner dyad was associated with both higher disclosure acceptability and lower co-owner misalignment, whereas our exploratory four-item empathy measure was not. In contrast, AI models significantly underperform human co-owners in anticipating the data acceptability, even when provided with within-dyad examples. These findings underscore a core HCI design challenge to develop privacy-aware AI that respects multi-stakeholder information boundaries.
Comments22 pages, 11 figures, 5 tables. Preprint
