Human-Centered Privacy Research in the Age of Large Language Models

• DOI: 10.48550/arxiv.2402.01994
• Publication Date: 2024-02-02
• Authors: Tianshi Li; Sauvik Das; Hao-Ping Lee; Dakuo Wang; Bingsheng Yao; Zhiping Zhang

The emergence of large language models (LLMs), and their increased use in user-facing systems, has led to substantial privacy concerns. To date, research on these concerns been model-centered: exploring how LLMs lead risks like memorization, or can be used infer personal characteristics about people from content. We argue that there is a need for more focusing the human aspect issues: e.g., design paradigms affect users' disclosure behaviors, mental preferences controls, tools, artifacts empower end-users reclaim ownership over data. build usable, efficient, privacy-friendly systems powered by with imperfect properties, our goal initiate discussions outline an agenda conducting human-centered issues LLM-powered systems. This Special Interest Group (SIG) aims bring together researchers backgrounds usable security privacy, human-AI collaboration, NLP, any other related domains share perspectives experiences this problem, help community establish collective understanding challenges, opportunities, methods, strategies collaborate outside HCI.