Friday, 24 March, 2023

Common AI Language Models Show Bias Against People With Disabilities

<p>A study was conducted at Penn State College and the researchers discovered considerable implicit bias against persons with impairments in all of the algorithms and models they examined. Previous research on pre-trained language models, which are trained on enormous quantities of data, has discovered sociodemographic biases against genders and ethnicities, but comparable biases against persons with disabilities have not been widely investigated until now.</p>