Understanding Users' Privacy Perceptions Towards LLM's RAG-based Memory

TL;DR

This study explores users' mental models, privacy concerns, and expectations regarding LLM's RAG-based memory, highlighting the need for transparency, control, and user-centric design in future memory systems.

Contribution

It provides an in-depth thematic analysis of user perceptions and practices related to LLM memory, revealing gaps in understanding and emphasizing privacy and control considerations.

Findings

Users have diverse, often incomplete mental models of LLM memory.

Privacy and control are major concerns among users.

Users want transparent, granular control over memory management.

Abstract

Large Language Models (LLMs) are increasingly integrating memory functionalities to provide personalized and context-aware interactions. However, user understanding, practices and expectations regarding these memory systems are not yet well understood. This paper presents a thematic analysis of semi-structured interviews with 18 users to explore their mental models of LLM's Retrieval Augmented Generation (RAG)-based memory, current usage practices, perceived benefits and drawbacks, privacy concerns and expectations for future memory systems. Our findings reveal diverse and often incomplete mental models of how memory operates. While users appreciate the potential for enhanced personalization and efficiency, significant concerns exist regarding privacy, control and the accuracy of remembered information. Users express a desire for granular control over memory generation, management, usage and updating, including clear mechanisms for reviewing, editing, deleting and categorizing memories, as well as transparent insight into how memories and inferred information are used. We discuss design implications for creating more user-centric, transparent, and trustworthy LLM memory systems.