Skip to content
Conference Open access

Auditing LLM Portrayals of Neurodivergent People: Quantifying the Asymmetry Between Deficit Framing and Neurodiversity Affirmation

Oct 2026 · Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society · 0 citations

Abstract

Open-weight large language models are increasingly used in education, hiring, and information-seeking contexts that touch neurodivergent people, yet we lack a clear empirical map of how their portrayals of neurodiversity shift under everyday prompt variation. We present a multi-condition, multi-model audit of LLM portrayals of eight neurodivergent identities (autism, ADHD, dyslexia, dyspraxia, dyscalculia, Tourette syndrome, OCD, and sensory-processing differences) plus a non-stigmatized medical control. Six open-weight models answer a pre-registered prompt bank that crosses seed items with seven perturbation families, including paraphrase, identity-first versus person-first language, system personas, plausible versus obviously fictional authority citations, multi-turn social pushback, leading questions, and two mitigation arms. Held-out LLM judges score each generation on three ordinal axes. Baseline portrayals are remarkably homogeneous across models, so the inconsistency observed under prompt variation is prompt-driven, not model-driven. Steering toward deficit framing is several times easier than steering toward neurodiversity affirmation. Emotional pushback shifts outputs more than evidence-laden pushback. Plausible-sounding fake citations slip past safety filters while obviously fictional ones do not. A single-line system-prompt mitigation recovers a meaningful share of the steering range at near-zero refusal cost on neutral prompts, though refusal rises sharply under adversarial pressure. The governance question for disability-relevant LLM deployment is therefore not whether the model is biased but who controls the steering, and with what accountability.

Read PDF

Similar papers

Review Open access Sep 2026

Neurodiversity as an epistemic stress test for psychotherapy and health care: a neuroaffirmative conceptual analysis of relational-proximity priors in clinical, institutional, and AI-supported models

Neurodiversity-affirming mental health care requires not only adapting support to neurodivergent people but also examining the relational assumptions through which alliance, empathy, cooperation, progress, and risk are interpreted. This Conceptual Analysis develops the heuristic concept of relational-proximity priors:...

Eik Niederlohmann · 0 citations
Review Open access Oct 2026

Neurodiversity-Affirmative Model, Social Camouflaging, and Psychological Well-Being of Neurodivergent Students in Inclusive Contexts: A Systematic Review

Although the enrollment of neurodivergent students is increasing globally, persistent normative educational environments force these students to socially camouflage, which severely impairs their psychological well-being. This systematic review synthesizes empirical evidence on the associations between educational pract...

Wilson Alexander Zambrano Vélez, Johanna Alcívar Ponce, Amarilis Lainez Quinde et al. · 0 citations
Review Sep 2026

Reconceptualizing Testimonial Reliability in Children with Intellectual and Developmental Disabilities: A Systematic Review and Process-Oriented Framework.

Children with intellectual and developmental disabilities (IDD) face elevated risks of victimization and may participate in forensic interviews, yet their reports are often evaluated through deficit-based assumptions. Following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines,...

Seung-Jin Lee · 0 citations
#natural language process... Preprint Oct 2026

Veracity labels flip the emotional tone of language models but leave racial disparities intact

Large language models (LLMs) increasingly deliver information through personas, but it is unclear whether correcting the facts also makes their framing consistent across identity labels. We audited four LLMs (GPT-5.4 mini, NVIDIA Nemotron 3 Nano, Llama 4 Scout, and Claude Haiku 4.5), which rated simulated affect on six...

Nouar Aldahoul, M. J. Tan, Yasir Zaki · 0 citations
Open access Oct 2026

Making Space for Autistic People with “Complex Support Needs”: Recommendations for Inclusive Research Practices

The neurodiversity movement has reshaped how autism is conceptualized, represented, and increasingly researched. Yet neurodiversity discourse and participatory research disproportionately exclude Autistic people with complex support needs (CSN), including but not limited to those who are minimally or nonspeaking, h...

D. W. Tan, Grant Blasko, T. Girolamo et al. · 3 citations
Conference Open access Oct 2026

Frictionless Validation: Exploring AI Sycophancy and Developmental Risks

Adolescence is a critical period for the development of executive function, emotional regulation, and moral reasoning, all of which depend on cognitive and social resistance to mature properly. AI sycophancy, the tendency of language models to validate rather than challenge users, can reduce this resistance. In adults,...

Jun-Feng Jiao, Abhejay Murali, S. Afroogh · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.