Do LLMs behave differently when the prompter is human vs another LLM?

by Dang Nguyen4 months ago
50

It seems clear that LLMs can recognize when a text is from a human or an LLM. We could make a reasonable assumption that they "expect" prompts to come from humans, have human style. What happens if the prompt, controlled for content, reads very LLM? Might they behave differently?

Implemented:https://github.com/Hypogenic-AI/llm-human-vs-llm-54c9-claudeImplemented:https://github.com/Hypogenic-AI/llm-human-vs-llm-ffea-gemini
Chat

If you are inspired by this idea, you can reach out to the authors for collaboration or cite it:

@misc{nguyen-do-llms-behave-2026,
  author = {Nguyen, Dang},
  title = {Do LLMs behave differently when the prompter is human vs another LLM?},
  year = {2026},
  url = {https://hypogenic.ai/ideahub/idea/dyWMLVUlyLEPhuzpQ2Ob}
}

Comments (0)

Please sign in to comment on this idea.

No comments yet. Be the first to share your thoughts!