EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Occupational Gender Bias in Chinese Generative AI Models: Cross-Model Evidence of Stereotypical Amplification and Systematic Underrepresentation

Yunhong Liu, Aijun Lin, Sui Peng, Zelong Cai

Systems · 09.03.2026 · 14 (3) · 286

DOI: 10.3390/systems14030286 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Systems, online erschienen am 09.03.2026. Verfasst von Yunhong Liu und 3 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 51 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Major Project of the National Social Science Fund of China, titled “Journalism Ethics and Regulation in the Age of Artificial Intelligence”.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Occupational gender stereotypes are widely embedded in social cognition and increasingly reproduced through generative artificial intelligence (AI). Two mainstream Chinese generative AI models (DeepSeek V3 and Qwen 2.5) were audited by eliciting occupation–gender pronoun associations for 72 census-anchored occupations using a standardized questionnaire and an automated testing pipeline. Each occupation was queried in 1000 independent rounds, yielding 2,880,000 item-level observations. The results show that, for both models, the fitted relationship between census female shares and model-implied female pronoun associations follows an S-shaped pattern. This pattern is consistent with a dominance-amplifying mapping that pushes male-dominated occupations toward lower female attribution and female-dominated occupations toward higher female attribution. Meanwhile, women’s overall visibility is consistently shifted downward: when the census benchmark is 50% female, the predicted female proportion remains below parity at 48% in DeepSeek and 43% in Qwen. Cross-model comparisons reveal substantial heterogeneity in bias profiles: DeepSeek primarily compresses female attribution in male-dominated occupations, whereas Qwen amplifies female dominance in occupations where women already predominate. Overall, these findings characterize a multi-layered output-level bias pattern combining structural amplification with a system-wide downward shift in women’s aggregate visibility.

Förderung

  • Major Project of the National Social Science Fund of China, titled “Journalism Ethics and Regulation in the Age of Artificial Intelligence”

Zitieren

APA

Liu, Y., Lin, A., Peng, S., & Cai, Z. (2026). Occupational Gender Bias in Chinese Generative AI Models: Cross-Model Evidence of Stereotypical Amplification and Systematic Underrepresentation. Systems, 14(3), 286. https://doi.org/10.3390/systems14030286

Vancouver

Liu Y, Lin A, Peng S, Cai Z. Occupational Gender Bias in Chinese Generative AI Models: Cross-Model Evidence of Stereotypical Amplification and Systematic Underrepresentation. Systems. 2026;14(3):286. doi:10.3390/systems14030286

BibTeX
@article{Liu2026,
  author = {Liu, Yunhong and Lin, Aijun and Peng, Sui and Cai, Zelong},
  title = {Occupational Gender Bias in Chinese Generative AI Models: Cross-Model Evidence of Stereotypical Amplification and Systematic Underrepresentation},
  journal = {Systems},
  year = {2026},
  volume = {14},
  number = {3},
  pages = {286},
  doi = {10.3390/systems14030286}
}
Kostenlos starten →