Abstract
Large Language Models (LLMs) are capable of generalizing human language for the completion of never-before-seen tasks, leading to widespread deployment. While this automation provides clear utility, completing these tasks often requires the insertion of Personally Identifiable Information (PII), strings of information that uniquely identify some individual, raising privacy concerns. However, ethics has prevented the curation of a public, authentic dataset of PII. Without an appropriate dataset, it is difficult to quantify privacy risks. Thus, we introduce the PANOPTICON pipeline and dataset.
The dataset, generated by Meta's Llama-3.1-8B-Instruct model, contains 67,718 prompts, intended for the model's context window, containing PII spans derived from 9,674 publicly available synthetic user profiles. We measure lexical diversity and S-BERT diversity of the created dataset to evaluate realism. Finally, we present a case study showcasing the utility of PANOPTICON data for understanding Prompt Inversion Attacks (PIAs). PANOPTICON thus emerges as the first benchmark dataset for studying PIAs over private corpora, providing a foundation for future LLM privacy research.
Blogger's Review: The introduction of the PANOPTICON dataset fills a crucial gap in LLM privacy research. By providing real PII examples, it lays the groundwork for evaluating model performance in the face of privacy attacks, potentially influencing privacy management strategies in practical applications. This research is significant for academia and offers guidance for real-world privacy concerns.