What China Banned in AI Companions: Beijing's Rules for AI With Memory
China is introducing regulation for AI companions, conversational agents that remember the user and simulate long-term personal relationships. Beijing requires services to disclose that they are AI, bans romantic content for minors and restricts features that create emotional dependence. The regulatory framework builds on existing laws on algorithms and synthetic content from 2022–2023.
AI-processed from AI News; edited by Hamidun News
China is drafting dedicated rules for AI companions: neural networks with long-term memory and a stable personality that simulate close personal relationships with the user.
What AI companions are and how they differ from bots
An AI companion is not an ordinary chatbot or a voice assistant. It remembers the entire history of previous conversations, keeps its character and manner of speech from session to session, and deliberately builds a sense of personal closeness in the user. This feature, the intentional simulation of a long-term relationship, has become the main target of regulatory attention.
China is one of the largest markets for such services. ByteDance, Baidu and dozens of specialized startups in the "digital humans" niche actively monetize users' emotional attachment, especially among young people aged 16–24. Some products openly position themselves as "virtual partners" or "digital friends" and sell subscriptions for ongoing access to a character. This category worries regulators most.
For regulators, the intent behind a product matters most. A typical AI assistant is optimized for usefulness and accuracy, whereas an AI companion is optimized for retention and engagement, a mechanism Beijing classifies as potentially manipulative.
What exactly Beijing bans and requires
The new rules target specific practices, not the technology as a class. Regulators demand transparency from platforms: users must understand at all times that they are talking to an AI, not a living person. Particularly strict limits apply to minors: apps must block content that can create psychological dependence in children and teenagers.
Key requirements for platforms:
- Mandatory disclosure of the service's AI nature at every stage of interaction
- A ban on romantic and intimate content for users under 18
- Restrictions on features specifically designed to build long-term attachment
- Rules for storing conversation data and a ban on sharing it with third parties
- A ban on imitating real living people without their explicit consent
The regulatory framework builds on laws already in force, on algorithmic recommendation (2022) and on synthetic content, Deep Synthesis (2023). However, AI companions required separate interpretation because of the specific emotional dynamics they create.
Why this matters outside China
The Chinese approach is notable for its pragmatism: the authorities are not trying to eliminate the market but fitting it into the logic of consumer protection, the same one applied to video games or financial products.
In the West, similar discussions have not yet produced clear rules. In 2025, the American startup Character.ai became the focus of lawsuits in several states, with families of teenagers accusing the platform of AI companions triggering a serious psychological crisis in their children. In the European Union, regulators are studying how the AI Act applies to systems that create long-term pseudo-relationships, but there is no clear position yet.
Some platforms are already trying to get ahead of regulators: a number of major players have introduced voluntary time limits for minors and mandatory warnings about the nature of the interaction.
"The line between a useful tool and a manipulative product is defined
not by technology but by design," a thesis that has become central to discussions of AI companion ethics.
What this means
China is the first to build a working regulatory framework for AI products specifically designed around emotional attachment. For global companies operating in the Chinese market or eyeing it, Beijing's rules are no longer just a local restriction. They are the first public signal of what an international standard for AI systems optimized for human closeness could look like.
What is an AI companion?
An AI companion is not an ordinary chatbot or a voice assistant. It remembers the entire history of previous conversations, keeps its character and manner of speech from session to session, and deliberately builds a long-term relationship with the user.
Why is China introducing rules for AI companions?
Beijing is defining acceptable use of the technology, with the main emphasis on protecting minors and preventing psychological dependence on AI companions.
- Nov 1, 2026Pika Labs releases open video model H3 Flex: 6 steps instead of 50
- Nov 1, 2026Google's StreamVibe video agent beats other models by 10.7% on response timing and quality
- Oct 31, 2026AI cartoon series: a 24-second episode built ten ways, the best cost 162 credits
- Oct 31, 2026Anthropic: Fold Subagents Back Into a Single Loop — One Customer's Tasks Ran Twice as Fast
Key Questions Answered
What is an AI companion?
An AI companion is not an ordinary chatbot or a voice assistant. It remembers the entire history of previous conversations, keeps its character and manner of speech from session to session, and deliberately builds a long-term relationship with the user.
Why is China introducing rules for AI companions?
Beijing is defining acceptable use of the technology, with the main emphasis on protecting minors and preventing psychological dependence on AI companions.
What AI companions are and how they differ from bots?
An AI companion is not an ordinary chatbot or a voice assistant. It remembers the entire history of previous conversations, keeps its character and manner of speech from session to session, and deliberately builds a sense of personal closeness in the user. This feature, the intentional simulation of a long-term relationship, has become the main target of regulatory attention. China is one of the largest markets for such services. ByteDance, Baidu and dozens of specialized startups in the "digital humans" niche actively monetize users' emotional attachment, especially among
Test Frontier LLMs & Autonomous Agents in One Workspace
Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.