China’s ChatGPT-like Service Adapts to Censorship Regime: 360 Highlights

Chinese tech firm 360 Security Technology has revealed its chatbot technology called Zhinao, or “intelligent brain’’, that is integrated with a multilevel filtering and moderation system to comply with the Chinese Communist Party’s strict censorship policies. The chatbot ends conversations immediately if users input words or phrases banned by the government. These words or phrases are routinely updated by human moderators and the public security department. Recently, Chinese authorities drafted new rules on generative AI, requiring a security assessment of each product before public release, and banning chatbots from generating content that suggests regime subversion, violence, pornography, or anything that “disrupts economic or social order”. Zhinao, the latest effort by Chinese firms to replicate OpenAI’s ChatGPT, has adopted the restriction policy common among the generative AI product offerings from major Chinese tech companies.