aiqbench analysis
Research reveals that attempts to enforce neutrality in LLM-generated narratives often result in the erasure of feminine perspectives, with female anthropomorphic characters appearing in only 2.2% of stories compared to 40.6% for masculine ones. While models frequently default to gender-neutral terms or avoid assignment entirely, this strategy inadvertently reinforces a masculine bias rather than achieving true fairness. The study suggests that future bias mitigation strategies must move beyond simple neutrality to actively distribute social possibilities more equally across imagined subjects.
Neutrality Bites: Gender Representation in AI-Generated Animal Stories
arXiv:2606.07969v1 Announce Type: new Abstract: Gender bias in AI-generated stories is a well-documented problem. While much attention has been paid to reducing or mitigating this bias, it is not always clear whether interventions produce genuinely fairer results. To investigate this issue, we examine how large language models (LLMs) handle gender assignment in a narrative context that is popular, highly ambiguous, and also known to closely reproduce human stereotypes: stories about talking animals. We prompt six leading LLMs to complete an English-language story about seven different anthropomorphic animal characters whose gender is unstated. We additionally iterate with four different narrative settings and a range of model temperatures. Across the 23.8K stories, we find that models frequently avoid gendering the animal character in the story (19% on average) or use gender-neutral language like "it" or "its" (38.2% on average). However, when gender is assigned, there is a significant masculine bias. Feminine animal characters are virtually absent, present in just 2.2% of stories vs. 40.6% that feature masculine characters. Our findings point to a broader argument: neutrality bites. In other words, models that prioritize neutrality to address social bias may actually contribute to the erasure of marginalized perspectives and identities. We suggest that alternative strategies beyond neutrality need to be pursued, such as ones that more equally distribute social possibilities across imagined subjects.
Abstract:Gender bias in AI-generated stories is a well-documented problem. While much attention has been paid to reducing or mitigating this bias, it is not always clear whether interventions produce genuinely fairer results. To investigate this issue, we examine how large language models (LLMs) handle gender assignment in a narrative context that is popular, highly ambiguous, and also known to closely reproduce human stereotypes: stories about talking animals. We prompt six leading LLMs to complete an English-language story about seven different anthropomorphic animal characters whose gender is unstated. We additionally iterate with four different narrative settings and a range of model temperatures. Across the 23.8K stories, we find that models frequently avoid gendering the animal character in the story (19% on average) or use gender-neutral language like "it" or "its" (38.2% on average). However, when gender is assigned, there is a significant masculine bias. Feminine animal characters are virtually absent, present in just 2.2% of stories vs. 40.6% that feature masculine characters. Our findings point to a broader argument: neutrality bites. In other words, models that prioritize neutrality to address social bias may actually contribute to the erasure of marginalized perspectives and identities. We suggest that alternative strategies beyond neutrality need to be pursued, such as ones that more equally distribute social possibilities across imagined subjects.
Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
arXiv:2606.07969 [cs.CL]
(or arXiv:2606.07969v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2606.07969 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
https://doi.org/10.1145/3805689.3812287 Focus to learn more DOI(s) linking to related resources
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 323 words
Related coverage
The J-Space Debate, Agent Swarms, and Pacing Frontier AI - Digital Minds Newsletter #4
LessWrong · Sep 18, 2026
I Did This Without Your Knowledge
LessWrong · Oct 17, 2026
Alignment Reduces Expressed but Not Encoded Gender Bias: A Unified Framework and Study
arXiv · Sep 11, 2026
Personality Shapes Gender Bias in Persona-Conditioned LLM Narratives Across English and Hindi: An Empirical Investigation
arXiv · Jun 5, 2026