How Chinese AI Chatbots Censor Themselves
Wired Business
•
Generative AI
Researchers from Stanford and Princeton found that Chinese AI models are likely than their Western counterparts to dodge political questions or deliver inaccurate answers.