LLMs Can Reveal the Identities of Pseudonymous Users at Scale with Remarkable Precision
Researchers Highlight Risks of LLM-Based Deanonymization Techniques Recent studies have illuminated the growing ability of large language models (LLMs) to deanonymize users online, posing significant privacy concerns for various stakeholders. In an experiment involving the Netflix dataset, researchers evaluated the…