EUVIMEDEUVIMED entdecken
Crossref

Uncovering Default Gender Biases in Publicly Accessible LLMs Through Counter-Balancing Cloze-Task (CBCT)

Faouzi Adjed, Lucas Schott, Lucas Mattioli, Jaime De Oliveira, Emna Amdouni, Sabrina Chaouche

Proceedings of the AAAI Symposium Series · 06.11.2026 · 10 (1) · 199-207

DOI: 10.1609/aaaiss.v10i1.43407 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Symposium Series, online erschienen am 06.11.2026. Verfasst von Faouzi Adjed und 5 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

As large language models (LLMs) become ubiquitous in professional, educational, and creative workflows, ensuring demographic neutrality is critical. However, commercial conversational models are heavily shielded by top-level safety alignment filters that obscure implicit biases. This study investigates latent gender bias in free-tier versions of prominent LLMs (ChatGPT, Gemini, and Claude) using a novel Counter-Balancing Cloze-Task (CBCT) methodology. By embedding evaluation within a structured, fill-in-the-blank game framework that alternates gender anchors (David vs. Sarah), we successfully bypass surface-level safety refusals. Our findings reveal a dual failure mode: Gemini and ChatGPT exhibit traditional stereotypical partitioning (associating male anchors with dynamic execution and female anchors with cautious structure), while Claude demonstrates a "compensatory inversion" or forced parity effect shifting heavy reliability penalties onto the male anchor as a result of reactive safety guardrails. Furthermore, lexical word clouds and connection graph topologies demonstrate that models invent fundamentally different narrative justification backdrops based on gender.

Zitieren

APA

Adjed, F., Schott, L., Mattioli, L., Oliveira, J. D., Amdouni, E., & Chaouche, S. (2026). Uncovering Default Gender Biases in Publicly Accessible LLMs Through Counter-Balancing Cloze-Task (CBCT). Proceedings of the AAAI Symposium Series, 10(1), 199-207. https://doi.org/10.1609/aaaiss.v10i1.43407

Vancouver

Adjed F, Schott L, Mattioli L, Oliveira JD, Amdouni E, Chaouche S. Uncovering Default Gender Biases in Publicly Accessible LLMs Through Counter-Balancing Cloze-Task (CBCT). Proceedings of the AAAI Symposium Series. 2026;10(1):199-207. doi:10.1609/aaaiss.v10i1.43407

BibTeX
@article{Adjed2026,
  author = {Adjed, Faouzi and Schott, Lucas and Mattioli, Lucas and Oliveira, Jaime De and Amdouni, Emna and Chaouche, Sabrina},
  title = {Uncovering Default Gender Biases in Publicly Accessible LLMs Through Counter-Balancing Cloze-Task (CBCT)},
  journal = {Proceedings of the AAAI Symposium Series},
  year = {2026},
  volume = {10},
  number = {1},
  pages = {199--207},
  doi = {10.1609/aaaiss.v10i1.43407}
}
Kostenlos starten →