@inproceedings{col-chan-2026-unintended,
    title = "Unintended Effects of Geographic Conditioning in Large Language Models",
    author = "Col, Naz  and
      Chan, David M.",
    editor = "Mysore, Sheshera  and
      Kumar, Sachin  and
      Balachandran, Vidhisha  and
      Hayati, Shirley Anugrah  and
      Brahman, Faeze  and
      Moussa, Hanane Nour  and
      Salemi, Alireza",
    booktitle = "Proceedings of the Second Workshop on Customizable {NLP}: Progress and Challenges in Customizing {NLP} for a Domain, Application, Group, or Individual ({C}ustom{NLP}4{U})",
    month = jul,
    year = "2026",
    address = "San Diego, California, USA",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl-workshops/2026.customnlp4u-1.18/",
    pages = "191--201",
    ISBN = "979-8-89176-396-8",
    abstract = "Modern conversational AI systems frequently rely on user metadata to localize responses, yet the unintended regional biases introduced by this hidden context remain poorly understood. In this work, we evaluate {\_}location leakage{\_}: the phenomenon where a model generates geographic references despite receiving a geographically neutral user prompt. Across both creative writing and open-ended Q{\&}A prompts, even state-of-the-art LLMs systematically favor region-specific outputs when exposed to location metadata, with leakage spiking by up to 793 times above baseline (e.g., from 0.04{\%} to 31.7{\%} for Llama 3.1-8B, and 21.3{\%} and 8.8{\%} for Qwen3-8B and Claude Sonnet 4.6, respectively). Our analysis further shows a novel structural conditioning effect: replacing the injected location with the placeholder ``Unknown'' still elevates leakage by up to 72 times above baseline, demonstrating that the user profile frame itself, independent of any geographic content, acts as a generative conditioning signal."
}