Beijing (AsiaNews) – China has introduced a series of rules to regulate artificial intelligence (AI) services designed to build emotional bonds with users.
New rules issued by the Cyberspace Administration of China, along with four other ministries, went into effect on 15 July, imposing limits on chatbots capable of simulating emotions, personalities, and human relationships. Minors are specifically banned from interacting with “virtual partners" or "digital relatives”.
The measures are part of Beijing's strategy to exert ever-tighter control over the development of generative AI, which is rapidly expanding in the country.
In recent years, millions of Chinese users have used apps that offer the ability to create virtual girlfriends and boyfriends inspired by movie and TV characters, digitally revive deceased family members, or transform chatbots into permanent confidants and best friends.
According to the new rules, platforms will not be able to generate content for minors that encourages dangerous behaviour, provokes extreme emotions, or promotes habits harmful to physical and mental health.
AI systems will also be required to avoid generating emotional dependence, compromising real-life relationships, or manipulating users into making unreasonable decisions.
Platforms will be required to send users exhibiting addictive behaviour an excessive usage reminder every two hours.
These restrictions apply only to services designed to foster emotional engagement between users and chatbots. Intelligent customer service, workplace assistants, or knowledge-sharing tools are excluded.
As Zhang Linghan, director of the Institute of Artificial Intelligence Law at the China University of Political Science and Law in Beijing, explained to China News Weekly, the emotional component offered by these anthropomorphic services can easily turn into psychological dependence, reducing users' ability to build deep relationships and manage conflicts in real life.
Chinese authorities have long denounced these risks, and the new provisions, in addition to seeking to protect minors, confirm Beijing's vision of governing AI development based on criteria of social stability and national security.
Major Chinese technology companies began to adapt even before the new regulations came into effect.
On 15 July, ByteDance, owner of TikTok, and e-commerce giant Alibaba disabled the virtual companion features integrated into their AI chatbots, Doubao and Qianwen, respectively, which are among the most widely used AI services in the country. Alibaba immediately discontinued the service, while ByteDance will allow users to transfer their data before 15 October.
Tencent, owner of the WeChat messaging platform and one of China's largest technology groups, also removed a similar feature from its Yuanbao chatbot at the end of June.
Over the past two years, several companies have already taken action following reports from the authorities about sexually explicit conversations.
In 2023, MiniMax, one of China's leading AI startups, withdrew its Glow chatbot from app stores. A company official told Chinese media that “80% of users would engage in suggestive or sexually explicit conversations with the AI companions they created within Glow.”
Last June, in Shanghai, the authorities reprimanded AI chat app Zhumengdao for producing vulgar and inappropriate content for minors.
Now the new rules could reshape the market. After Doubao and Qianwen were shut down, many users moved to competing platforms, many of which began requiring identity verification before granting access to their virtual assistants.
RED LANTERNS IS THE ASIANEWS NEWSLETTER DEDICATED TO CHINA. WOULD YOU LIKE TO RECEIVE IT EVERY THURSDAY? TO SUBSCRIBE, CLICK HERE.
