China deploys censors to create socialist AI

Chinese government officials are testing AI models to ensure they reflect socialist values, expanding censorship efforts. The Cyberspace Administration of China is reviewing AI models from tech giants like ByteDance and Alibaba, focusing on political sensitivities and President Xi Jinping-related content. Companies must filter out sensitive information and update keywords violating socialist values weekly. Chatbots reject queries on topics like Tiananmen Square or President Xi, with ByteDance's model excelling in compliance. Developers face challenges in controlling AI text output, resorting to real-time replacements for sensitive content. China aims to establish safety protocols for AI models, emphasizing real-time monitoring. The stringent approval process pushes AI groups to adapt quickly to censorship demands, with ByteDance leading in aligning with Beijing's narratives. The government limits the rejection of questions by AI models, promoting controlled responses on political topics. Chinese engineers navigate complexities to ensure AI generates politically correct answers, reflecting the country's regulatory regime on AI content.