Study of 68,000 Prompts Finds People Treat LLMs as Oracles

A Stanford-led preprint analyzing 68,000 prompts and 140,000 messages finds users increasingly ask LLMs to tell them what to think rather than help them think, with usage growing yearly—especially among younger users—raising risks to autonomy and well-being.

2026-09-16 ~ 2026-09-17 · 3 related posts

• Study: people increasingly treat LLMs as oracles, with younger users most reliant — stanfordnlp · 2026-09-16
• Study of 68,000 prompts finds users increasingly treat LLMs as oracles, often unaware — ValerioCapraro · 2026-09-17
• Full paper: LLM users increasingly ask models to think for them, often unaware — ValerioCapraro · 2026-09-17