LLM Disability Bias

Systematic prejudice against people with disabilities embedded in large language models due to biased training data and development processes that underrepresent disabled communities. Research has documented multiple forms of this bias: pretrained language models associate disability-related words with negative contexts, GPT-based resume screening ranks candidates with disability-related achievements lower than identical candidates without them, and chatbots generate stories that assume disabled people desire to be "fixed" or depend on non-disabled help. Addressing LLM disability bias requires both technical debiasing and centering disabled people in AI development.