Adultification Bias in LLMs and Text-to-Image Models

Jane Castleman, Aleksandra Korolova

The rapid adoption of generative AI models in domains such as education, policing, and social media raises significant concerns about potential bias and safety issues, particularly…