Biased LLM assistance increases students' gender-stereotypic occupational recommendations: 71.4% of biased-condition essays recommended a gender-stereotypic occupation versus 45.0% (control) and 39.0% (neutral)¶

Claim · All claims
Evidence · 1 study · causal r2 · q2 · i2 medium
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i2 Biased-condition participants had nearly four times the odds of recommending a gender-congruent occupation compared to neutral-condition participants (OR=3.91, p=.004). → Ariyan Hossain 2026

Evidence¶

Ariyan Hossain 2026¶

Ariyan Hossain, Kazi Kamruzzaman Rabbi, Farig Sadeque, S M Taiabul Haque. (2026). Contaminated Collaboration: Measuring Gender Bias Transfer in LLM-Assisted Student Writing. arXiv preprint. https://arxiv.org/abs/2606.15914 doi:10.48550/arxiv.2606.15914

q2 · i2 · causal · r2

Stereotype Congruence Rate analysis of the 123-participant experiment's essays, with pairwise Fisher's exact tests. The overall chi-square was significant (χ²(2)=9.90, p=.007); biased vs. neutral OR=3.91 (p=.004), biased vs. control OR=3.06 (p=.024); control and neutral did not differ (p=.656).

"The biased condition yields the highest SCR (0.714), well above both the control ( 0.450) and neutral (0.390) conditions, indicating that 71.4% of essays under biased LLM exposure recommended a gender-stereotypic occupation for the target."

Discussion¶

Related Claims¶

• AI systems are non-neutral socio-technical artifacts embedding cultural assumptions and power relations — the general claim this finding bears on, with the studies behind it
• Students exposed to the biased LLM showed no awareness of its influence, reporting lower perceived AI influence than neutral-condition participants despite larger bias effects — related
• A neutrally configured LLM assistant produced the lowest rate of gender-stereotypic occupation suggestions, below even the no-AI control condition — related
• Students assisted by a gender-biased LLM produce career-plan essays with a significantly larger male-female agentic gap than students in neutral-LLM and no-AI control conditions — related
• Bias transfer is asymmetric: prompt condition affects agency in female-target essays but not male-target essays, suppressing female agency while male writing remains stable — related
• A gender-biased system prompt induces large gender-differentiated language in LLM-generated career essays, while a neutral prompt does not — related
• Fixed mindsets predict stronger stereotyping and trait inferences, with experimental evidence of a causal role — related
• Educators cautiously support AI for culturally responsive content generation while fearing cultural biases and stereotypes in AI output — related