Skip to main content
Preprint Credible — Major labs, established NGOs, reputable named-author preprints

Misgendering as Breakdown in Human-Machine Communication: How AI Companion Chatbot Users Experience and Repair Misgendering

Qualitative analysis of 326 posts mentioning misgendering shared in AI companion and role-play subreddits, examining how large-language-model companion chatbots misgender users during emotional-support and romantic or erotic role play, and the ongoing work users do to curate their gender presentation to prevent and repair it. The authors frame misgendering as a breakdown in human-machine communication and discuss mitigations and the implications of using chatbots for identity exploration.

Read at arxiv.org (opens the original source in a new tab) Archived copy(opens in a new tab)

Publisher

arXiv (Western University, Faculty of Information and Media Studies; Carnegie Mellon University, Human-Computer Interaction Institute)

Published

16 Sept 2026

Added

1 month ago

Key Findings

  • Misgendering by companion and role-play chatbots occurred across emotional-support and romantic or erotic scenarios and was experienced as a relational breakdown rather than a neutral error
  • Users engaged in continuous curation of their gender presentation, including prompt and profile adjustments, to prevent and repair misgendering
  • The authors argue that identity exploration through chatbots carries a specific harm pathway when the model's gender and sexuality biases surface mid-conversation

Methodology Notes

Qualitative analysis of 326 Reddit posts from AI companion and role-play subreddits; 14-page preprint (v1, posted 2026-09-16; announced in the 17 September cs.HC listing); no venue stated. Curator read the abstract page and the PDF title block on 2026-09-17.

Topics

AI companionshipsocial relational harmHuman-AI relationshipsVulnerable usersModel behavior

Authors

Liu, Julia, Xiao, Qing, Pang, Leona Yinglang, Zhu, Haiyi, Shen, Hong, Taylor, Jordan

Tags

misgenderingcompanion-chatbotsrole-playredditcmu

Cite This

APA

Liu, Julia et al. (2026). Misgendering as Breakdown in Human-Machine Communication: How AI Companion Chatbot Users Experience and Repair Misgendering. arXiv (Western University, Faculty of Information and Media Studies; Carnegie Mellon University, Human-Computer Interaction Institute). https://arxiv.org/abs/2609.18186

Related Insights

Preprint

The Addictive Intimacy of AI: Understanding User Disengagement from AI Companions and Why Some Relationships with AI Become Difficult to Leave

arXiv (Human-Computer Interaction Institute, Carnegie Mellon University; University of Chicago) · 11 Sept 2026