arXiv subjects

Mahika Phutane

Publications and source records attributed to Mahika Phutane.

5 recordsLinked to original sources

arXiv2025

ABLEIST: Intersectional Disability Bias in LLM-Generated Hiring Scenarios

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.

cs.CL↗

arXiv2025

Disability Across Cultures: A Human-Centered Audit of Ableism in Western and Indic LLMs

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.

cs.CY↗

arXiv2025

"Cold, Calculated, and Condescending": How AI Identifies and Explains Ableism Compared to Disabled People

People with disabilities (PwD) regularly encounter ableist hate and microaggressions online. These spaces are generally moderated by machine learning models, but little is known about how effectively AI models identify ableist speech and how well their judgments align with PwD. To investigate this, we curated a first-of-its-kind dataset of 200 social media comments targeted towards PwD, and prompted state-of-the art AI models (i.e., Toxicity Classifiers, LLMs) to score toxicity and ableism for each comment, and explain their reasoning. Then, we recruited 190 participants to similarly rate and explain the harm, and evaluate LLM explanations. Our mixed-methods analysis highlighted a major disconnect: AI underestimated toxicity compared to PwD ratings, while its ableism assessments were sporadic and varied. Although LLMs identified some biases, its explanations were flawed--they lacked nuance, made incorrect assumptions, and appeared judgmental instead of educational. Going forward, we discuss challenges and opportunities in designing moderation systems for ableism, and advocate for the involvement of intersectional disabled perspectives in AI.

cs.HC↗

arXiv2024

"It's Kind of Context Dependent": Understanding Blind and Low Vision People's Video Accessibility Preferences Across Viewing Scenarios

While audio description (AD) is the standard approach for making videos accessible to blind and low vision (BLV) people, existing AD guidelines do not consider BLV users' varied preferences across viewing scenarios. These scenarios range from how-to videos on YouTube, where users seek to learn new skills, to historical dramas on Netflix, where a user's goal is entertainment. Additionally, the increase in video watching on mobile devices provides an opportunity to integrate nonverbal output modalities (e.g., audio cues, tactile elements, and visual enhancements). Through a formative survey and 15 semi-structured interviews, we identified BLV people's video accessibility preferences across diverse scenarios. For example, participants valued action and equipment details for how-to videos, tactile graphics for learning scenarios, and 3D models for fantastical content. We define a six-dimensional video accessibility design space to guide future innovation and discuss how to move from "one-size-fits-all" paradigms to scenario-specific approaches.

cs.HC↗

arXiv2022

Tactile Materials in Practice: Understanding the Experiences of Teachers of the Visually Impaired

Teachers of the visually impaired (TVIs) regularly present tactile materials (tactile graphics, 3D models, and real objects) to students with vision impairments. Researchers have been increasingly interested in designing tools to support the use of tactile materials, but we still lack an in-depth understanding of how tactile materials are created and used in practice today. To address this gap, we conducted interviews with 21 TVIs and a 3-week diary study with eight of them. We found that tactile materials were regularly used for academic as well as non-academic concepts like tactile literacy, motor ability, and spatial awareness. Real objects and 3D models served as "stepping stones" to tactile graphics and our participants preferred to teach with 3D models, despite finding them difficult to create, obtain, and modify. Use of certain materials also carried social implications; participants selected materials that fostered student independence and allow classroom inclusion. We contribute design considerations, encouraging future work on tactile materials to enable student and TVI co-creation, facilitate rapid prototyping, and promote movement and spatial awareness. To support future research in this area, our paper provides a fundamental understanding of current practices. We bridge these practices to established pedagogical approaches and highlight opportunities for growth regarding this important genre of educational materials.

cs.HC↗