Can LLMs Keep a Secret? Testing Privacy Implications of Language Models via Contextual Integrity Theory

Open in new window