LLM Agents Test Where AI Sharing Private Info Goes Too Far

Original: A Generative Model of Contextual Integrity: Appropriate vs. Inappropriate Information Sharing

Source: arxiv.org ↗