Internet

How AI bias can creep into online content moderation

A University of Queensland study has shown large language models (LLMs) used in AI content moderation may be prone to subtle biases that undermine their neutrality. A team led by data scientist Professor Gianluca Demartini ...

page 6 from 40