Model Hypnosis: Strong control of AI via additive subliminal effects

Researchers at MIT and Harvard, led by Enric Boix-Adserà, have demonstrated that AI models are broadly susceptible to 'model hypnosis,' where individually weak and seemingly irrelevant cues in prompts can be systematically combined to strongly control model behavior. The phenomenon occurs across model families and scales, including frontier reasoning models, and hypnotic prompts can transfer between models, posing new challenges for AI safety and interpretability.

Computer Science Computation and Language Submitted on 17 Aug 2026 Title:Model Hypnosis: Strong control of AI via additive subliminal effects View PDF /pdf/2608.16834 HTML experimental https://arxiv.org/html/2608.16834v1 Abstract:We demonstrate that AI models are broadly susceptible to a phenomenon we call model hypnosis, in which individually weak and seemingly irrelevant cues in the prompt can be systematically combined to strongly control model behavior. Model hypnosis occurs across model families and scales, including in frontier reasoning models, and hypnotic prompts can transfer between models. Because the model is controlled by inconspicuous textual choices, such as paraphrases and typos, model hypnosis presents new challenges and avenues for AI safety, and is a major hurdle for AI interpretability. Submission history From: Enric Boix-Adserà view email /show-email/801763f6/2608.16834 v1 Mon, 17 Aug 2026 17:20:10 UTC 4,688 KB References & Citations Loading... Bibliographic and Citation Tools Bibliographic Explorer What is the Explorer? https://info.arxiv.org/labs/showcase.html arxiv-bibliographic-explorer Connected Papers What is Connected Papers? https://www.connectedpapers.com/about Litmaps What is Litmaps? https://www.litmaps.co/ scite Smart Citations What are Smart Citations? https://www.scite.ai/ Code, Data and Media Associated with this Article alphaXiv What is alphaXiv? https://alphaxiv.org/ CatalyzeX Code Finder for Papers What is CatalyzeX? https://www.catalyzex.com DagsHub What is DagsHub? https://dagshub.com/ Gotit.pub What is GotitPub? http://gotit.pub/faq Hugging Face What is Huggingface? https://huggingface.co/huggingface ScienceCast What is ScienceCast? https://sciencecast.org/welcome Demos Recommenders and Search Tools Influence Flower What are Influence Flowers? https://influencemap.cmlab.dev/ CORE Recommender What is CORE? https://core.ac.uk/services/recommender arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs https://info.arxiv.org/labs/index.html .