AI bias detection tool promises to tackle discrimination in models

Generative AI models like ChatGPT are trained using vast amounts of data obtained from websites, forums, social media and other online sources; as a result, their responses can contain harmful or discriminatory biases.