China's Cyberspace Administration issued draft rules tightening oversight of AI that mimic human personalities; they're open for comment until Jan 25. They require AI identity disclosure, security reviews, intervention for addictive or extreme emotional use, and reports to provincial authorities when launching human-like features or exceeding user thresholds.

Chinese draft rules will likely impose immediate legal and compliance obligations on consumer-facing AI providers, forcing security and ethical reviews, user transparency, reporting to provincial regulators when human-like features launch or usage crosses thresholds, and intervention for addictive or extreme emotional use. These changes will reshape product design, operations and market conduct across China's AI sector.