China Tightens AI Personality Rules

China's cyber regulator has issued draft rules targeting artificial intelligence services that simulate human personalities and engage users emotionally. The proposed regulations underscore Beijing's commitment to shaping the development of consumer-facing AI by enforcing stricter safety and ethical requirements.

The new rules would apply to AI products and services publicly available in China that exhibit human-like personality traits and interact emotionally through various media. Service providers would be responsible for warning users about excessive usage and implementing interventions for signs of addiction.

Furthermore, providers must ensure algorithm safety, data security, and personal information protection throughout the product lifecycle. The draft also addresses psychological risks by requiring providers to assess user dependence and intervene in cases of extreme emotions or addictive behavior. Crucially, these AI services must adhere to content restrictions, prohibiting the generation of content that jeopardizes national security, spreads misinformation, or promotes violence and obscenity.