China Targets Emotional AI as ByteDance and Alibaba Revamp Chatbots
ByteDance Ltd. and Alibaba Group Holding Ltd. are discontinuing features that allowed users to create and interact with AI companions, in anticipation of forthcoming Chinese regulations concerning human engagement with artificial intelligence. ByteDance’s Doubao, China’s most popular AI chatbot, will discontinue a feature on July 15 that enables users to personalise their own AI personas, as indicated by an app notification. The notice instructed users to utilise a distinct, independent companion application. Alibaba’s Qwen has issued a similar alert, as have other major platforms like Tencent Holdings Ltd.’s Yuanbao, according to reports.
The abrupt withdrawal occurs in anticipation of forthcoming regulations from Beijing, set to be implemented in mid-July, aimed at enhancing oversight on humanlike AI services. Concerns regarding AI chatbots emulating human personalities and emotions, along with the potential attachments users may form to these interactions, prompted the introduction of new regulations, initially revealed in April. The new directives signify a notably extensive array of regulations aimed at mitigating the potential risks associated with AI, despite warnings from certain companies that stringent rules may hinder innovation. They indicate an increasing global concern regarding the psychological impact of conversational AI. In the United States, technology platforms are facing significant legal examination regarding comparable functionalities.
OpenAI Inc. and Character, backed by Alphabet Inc.AI has encountered a series of prominent lawsuits claiming that their hyper-realistic chatbots have cultivated perilous emotional dependencies and, in severe instances, contributed to the suicides of susceptible users. Beijing’s new regulations, propelled by the Cyberspace Administration of China, restrict platforms from producing content that incites extreme emotions in minors or cultivates unhealthy emotional dependencies that undermine real-world relationships. The framework additionally prohibits providers from utilising sensitive user conversation data for the training of future AI models. Chinese chatbot platforms have historically provided users with the capability to tailor an AI agent through a limited set of text prompts.
Among the most sought-after options are virtual companions, unlicensed digital therapists, and simulated replicas of popular music icons. Beyond AI chatbots, China’s examination of artificial intimacy is extending into tangible hardware. Two Chinese robotics industry groups are advocating for stricter ethical safeguards as a significant commercial influx of companion robots and full-size humanoids reaches consumers’ homes, according to a report.
