Harnessing LLMs for Scalable Data Minimization

Explore how Large Language Models can revolutionize data minimization practices in this 15-minute conference talk from PEPR '25. Discover how LLMs' advanced contextual understanding capabilities can accelerate data classification across organizational storage systems, enhance de-identification processes for text corpora, and streamline internal governance mechanisms compared to traditional approaches. Learn about proposed architectures that combine various LLM-based tools with complementary techniques like lineage tracing to enable proactive data minimization strategies and effectively prevent data sprawl across enterprise environments.