China targets human-like AI systems with new draft regulations

China has unveiled new draft rules, aiming to regulate AI with human-like interaction and tighten oversight of artificial intelligence specifically designed based on human personalities.

The recent move on the part of the Chinese cyber regulator highlights Beijing’s effort to govern AI and strengthen consumer-oriented safety and ethical requirements.

Key regulatory rules under new proposal

The draft delineates a regulatory approach which would ensure user well-being and addiction mitigation by warning them against excessive use.

Under the new draft, the service providers would be entrusted with safety responsibilities throughout the product lifecycle and identify the user state based on emotional well being.

The companies would also establish systems for algorithm review, data security, and personal information protection.

The proposal also targets psychological issues. The providers would be given responsibility to assess user states, their emotions and extent of dependency on the AI-powered service.

In the case of addictive attitude and extreme behavior, the companies would be entitled to take necessary measures of intervention.

The authorities also rolled out rules for content moderation, stating the services would be prohibited to generate content that threaten national security, spread misinformation or disinformation and promote violence and obscenity.

These newly proposed rules would be applied to all AI products and services offered to people in China. The AI systems that demonstrate human personality traits, communication styles, emotional interaction and thinking patterns, would be governed by these regulations.