> ## Content Index
> Fetch the complete content index at: https://kanon.ghost.io/llms.txt
> Use this file to discover other available public pages before exploring further.

# China's rules for human-like AI services are in force
- URL: https://kanon.ghost.io/chinas-rules-for-human-like-ai-services-are-in-force/
- Published: 2026-10-07T10:14:21.000Z
- Updated: 2026-10-07T10:14:21.000Z
- Description: AI companions and chatbots that simulate personality now face specific duties in China.
- Author: Kanon
- Tags: AI, China, Advertising, #Import 2026-09-30 14:47

**Background**

China's Interim Measures for the Administration of AI Anthropomorphic Interaction Services took effect on 15 July 2026, extending the regulatory framework to AI that simulates human personality and emotion.

**Three key points**

**1\. What the measures cover.** The rules apply to AI services that simulate human personality and emotion in interaction with users. This includes AI companions, conversational agents and branded chatbots that present as having a personality or emotional response. Services that answer factual questions without anthropomorphic characteristics are not the target.

**2\. Required safeguards.** Providers must implement AI identification prompts so users know they are interacting with an AI, anti-addiction controls, and protections for minors and older people. A safety assessment before launch is required, along with algorithm filing with the Cyberspace Administration.

**3\. Escalating enforcement powers.** Regulators can summon a provider's legal representative following a significant safety incident. Penalties escalate from warnings to service suspension and fines. Personal liability for the legal representative is built into the enforcement framework.

**Why this matters**

Brands using AI characters, virtual spokespeople or chatbots with a defined personality in China need to assess whether the measures apply. The combination of safety assessment, algorithm filing and identification prompts represents a compliance overhead specific to anthropomorphic use cases. Building these requirements into the product specification from the start is more efficient than retrofitting them.

**Sources:** [Just Security](https://www.justsecurity.org/148468/china-ai-companion-rules-relationships/?ref=kanon.ghost.io); [Deep Lex](https://www.deep-lex.com/ai-regulation-tracker/china?ref=kanon.ghost.io); [Regulations.AI](https://regulations.ai/regulations/RAI-CN-NA-SUMMARY-2026?ref=kanon.ghost.io).