analiza aiqbench
This study exposes critical safety blind spots in frontier LLMs by demonstrating how they amplify ableism and intersectional discrimination within high-stakes hiring scenarios. The authors introduce a novel evaluation framework that reveals state-of-the-art models often fail to detect subtle harms targeting candidates with intersecting marginalized identities like gender and caste. This work is essential for developers aiming to align AI systems with robust, inclusive safety standards beyond Western-centric benchmarks.
ABLEIST: Intersectional Disability Bias in LLM-Generated Hiring Scenarios
arXiv:2510.10998v2 Announce Type: replace Abstract: Large language models (LLMs) are increasingly under scrutiny for perpetuating identity-based discrimination in high-stakes domains such as hiring, particularly against people with disabilities (PwD). However, existing research remains largely Western-centric, overlooking how intersecting forms of marginalization--such as gender and caste--shape experiences of PwD in the Global South. We conduct a comprehensive audit of six LLMs across 2,820 hiring scenarios spanning diverse disability, gender, nationality, and caste profiles. To capture subtle intersectional harms and biases, we introduce ABLEIST (Ableism, Inspiration, Superhumanization, and Tokenism), a set of five ableism-specific and three intersectional harm metrics grounded in disability studies literature. Our results reveal significant increases in ABLEIST harms towards disabled candidates--harms that many state-of-the-art models failed to detect. These harms were further amplified by sharp increases in intersectional harms (e.g., Tokenism) for gender and caste-marginalized disabled candidates, highlighting critical blind spots in current safety tools and the need for intersectional safety evaluations of frontier models in high-stakes domains like hiring.
Abstract:Large language models (LLMs) are increasingly under scrutiny for perpetuating identity-based discrimination in high-stakes domains such as hiring, particularly against people with disabilities (PwD). However, existing research remains largely Western-centric, overlooking how intersecting forms of marginalization--such as gender and caste--shape experiences of PwD in the Global South. We conduct a comprehensive audit of six LLMs across 2,820 hiring scenarios spanning diverse disability, gender, nationality, and caste profiles. To capture subtle intersectional harms and biases, we introduce ABLEIST (Ableism, Inspiration, Superhumanization, and Tokenism), a set of five ableism-specific and three intersectional harm metrics grounded in disability studies literature. Our results reveal significant increases in ABLEIST harms towards disabled candidates--harms that many state-of-the-art models failed to detect. These harms were further amplified by sharp increases in intersectional harms (e.g., Tokenism) for gender and caste-marginalized disabled candidates, highlighting critical blind spots in current safety tools and the need for intersectional safety evaluations of frontier models in high-stakes domains like hiring.
To appear at AIES 2026. Please cite AIES version when the proceedings are available
Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Human-Computer Interaction (cs.HC); Machine Learning (cs.LG)
arXiv:2510.10998 [cs.CL]
(or arXiv:2510.10998v2 [cs.CL] for this version)
Text integral sub licența deschisă a publicației · 196 cuvinte
Subiecte conexe
How bad would it be if GPS satellites were shot down?
LessWrong · 12 iun. 2026
Amazon Research Awards recipients announced
Amazon Science · 27 mai 2026
Digital States 2026: AI Moves From Experiment to Infrastructure
govtech.com · 24 sept. 2026
Dyna-2: General Purpose Physical Agents with Hour-Long Autonomy — Jason Ma, Dyna Robotics
Latent Space · 13 oct. 2026