Ivoline Ngong

Ivoline Ngong, Swanand Kadhe, Hao Wang et al.

Conversational agents are increasingly woven into individuals' personal lives, yet users often underestimate the privacy risks associated with them. The moment users share information with these agents-such as large language models (LLMs)-their private information becomes vulnerable to exposure. In this paper, we characterize the notion of contextual privacy for user interactions with LLM-based Conversational Agents (LCAs). It aims to minimize privacy risks by ensuring that users (sender) disclose only information that is both relevant and necessary for achieving their intended goals when interacting with LCAs (untrusted receivers). Through a formative design user study, we observe how even "privacy-conscious" users inadvertently reveal sensitive information through indirect disclosures. Based on insights from this study, we propose a locally deployable framework that operates between users and LCAs, identifying and reformulating out-of-context information in user prompts. Our evaluation using examples from ShareGPT shows that lightweight models can effectively implement this framework, achieving strong gains in contextual privacy while preserving the user's intended interaction goals. Notably, about 76% of participants in our human evaluation preferred the reformulated prompts over the original ones, validating the usability and effectiveness of contextual privacy in our proposed framework. We opensource the code at https://github.com/IBM/contextual-privacy-LLM.

Tamunotonye Harry, Ivoline Ngong, Chima Nweke et al.

User interactions with language models vary due to static properties of the user (trait) and the specific context of the interaction (state). However, existing persona datasets (like PersonaChat, PANDORA etc.) capture only trait, and ignore the impact of state. We introduce Chameleon, a dataset of 5,001 contextual psychological profiles from 1,667 Reddit users, each measured across multiple contexts. Using the Chameleon dataset, we present three key findings. First, inspired by Latent State-Trait theory, we decompose variance and find that 74\% is within-person(state) while only 26\% is between-person (trait). Second, we find that LLMs are state-blind: they focus on trait only, and produce similar responses regardless of state. Third, we find that reward models react to user state, but inconsistently: different models favor or penalize the same users in opposite directions. We release Chameleon to support research on affective computing, personalized dialogue, and RLHF alignment.