BEIJING, China: China is moving to place new guardrails around artificial intelligence systems that behave like people, publishing draft regulations aimed at tightening control over AI services designed to simulate human personalities and form emotional connections with users.

The proposed rules, released by China's cyber regulator for public consultation, highlight Beijing's determination to guide the rapid expansion of consumer-facing AI with stronger safety, ethical, and behavioural standards.

Under the draft framework, the measures would apply to AI products and services offered to the public in China that present simulated human personality traits, thinking patterns, or communication styles. The rules cover systems that interact emotionally with users through text, images, audio, video, or other formats.

A central focus of the proposal is the risk of excessive use and emotional dependence. Service providers would be required to warn users about overuse and to intervene when signs of addiction emerge. The draft specifies that companies must monitor user behaviour and step in if interactions become unhealthy.

The proposed regulations would also impose lifecycle accountability on AI providers. Companies would be expected to assume responsibility for safety from development through deployment and to establish mechanisms for algorithm reviews, data security, and the protection of personal information.

Psychological risks are addressed directly in the draft. Providers would need to identify user states, assess emotional responses, and evaluate levels of reliance on the AI service. If users display extreme emotions or addictive behaviour, companies would be required to take "necessary measures" to intervene.

In addition to user well-being, the draft sets out explicit boundaries on content and conduct. AI services must not generate material that endangers national security, spreads rumours or promotes violence or obscenity, reinforcing long-standing regulatory red lines for digital content in China.