Disability Across Cultures: A Human-Centered Audit of Ableism in Western and Indic LLMs
Preprint
Abstract
People with disabilities (PwD) experience disproportionately high levels of discrimination and hate online, particularly in India, where entrenched stigma and limited resources intensify these challenges. Large language models (LLMs) are increasingly used to identify and mitigate online hate, yet most research on online ableism focuses on Western audiences with Western AI models. Are these models adequately equipped to recognize ableist harm in non-Western places like India? Do localized, Indic language models perform better? To investigate, we adopted and translated a publicly available ableist speech dataset to Hindi, and prompted eight LLMs--four developed in the U.S. (GPT-4, Gemini, Claude, Llama) and four in India (Krutrim, Nanda, Gajendra, Airavata)--to score and explain ableism. In parallel, we recruited 175 PwD from both the U.S. and India to perform the same task, revealing stark differences between groups. Western LLMs consistently overestimated ableist harm, while Indic LLMs underestimated it. Even more concerning, all LLMs were more tolerant of ableism when it was expressed in Hindi and asserted Western framings of ableist harm. In contrast, Indian PwD interpreted harm through intention, relationality, and resilience--emphasizing a desire to inform and educate perpetrators. This work provides groundwork for global, inclusive standards of ableism, demonstrating the need to center local disability experiences in the design and evaluation of AI systems.Keywords
All Related Versions
This publication has 34 references indexed in Scilit:
- Cultural bias and cultural alignment of large language modelsPNAS Nexus, 2024
- The methodology of studying fairness perceptions in Artificial Intelligence: Contrasting CHI and FAccTInternational Journal of Human-Computer Studies, 2022
- Intersectional discrimination against women and girls with disabilities in educational opportunities in IndiaWorld Development Perspectives, 2022
- Workplace disability inclusion in India: review and directionsManagement Research Review, 2020
- Agency of Autistic Children in Technology Research—A Critical Literature ReviewACM Transactions on Computer-Human Interaction, 2019
- Rhetoric of Online Disability Activism: #CripTheVote and Civic ParticipationCommunication, Culture & Critique, 2018
- The Coding Manual for Qualitative Researchers (3rd edition)Qualitative Research in Organizations and Management: An International Journal, 2017
- A new screening instrument for disability in low-income and middle-income settings: application at the Iganga-Mayuge Demographic Surveillance System (IM-DSS), UgandaBMJ Open, 2014
- Honorifics: A sociocultural verb agreement cue in Japanese sentence processingApplied Psycholinguistics, 2010
- Optimal number of response categories in rating scales: reliability, validity, discriminating power, and respondent preferencesActa Psychologica, 2000