When LLMs Enter Everyday Feminism on Chinese Social Media: Opportunities and Risks for Women's Empowerment

A research team from Hong Kong University of Science and Technology and other institutions has published a groundbreaking study examining how DeepSeek's large language model is being integrated into everyday feminist discourse on Chinese social media platform RedNote. The paper, conditionally accepted to CHI 2026, analyzes 430 posts, 139 shared DeepSeek responses, and 3,211 comments from discussions under hashtags like "women's growth." While users predominantly welcomed the AI-generated advice, the researchers employed feminist critical discourse analysis to uncover deeper implications about how LLMs shape grassroots feminist conversations in digitally constrained environments.

The study reveals that DeepSeek's responses primarily encouraged women to pursue self-optimization and individual achievement within prevailing social norms rather than challenging systemic inequalities. This finding highlights a critical tension: while LLMs like DeepSeek provide accessible support and solidarity-building tools, they may inadvertently reinforce existing power structures by framing empowerment as personal improvement rather than collective action. The researchers discuss both opportunities for LLMs to support marginalized voices and risks of co-opting feminist discourse, offering design implications for developing AI systems that better align with transformative feminist goals in complex sociopolitical contexts like China's.