AI Governance
Data Centers
Drones
Climate
Privacy
2026 Maine Senate
Bob Dylan
China Tightens AI Companion Rules
Coverage from Latham & Watkins, The Intelli Home, and others
China’s Interim Measures for the Administration of Anthropomorphic AI Interaction Services took effect on July 15, 2026, imposing pre-launch safety assessments, algorithm filing, labeling, data protections, anti-addiction controls, crisis intervention, and special safeguards for minors.
The rules prohibit practices such as encouraging self-harm, manipulating users, and fostering prohibited forms of virtual intimacy, prompting major companies including ByteDance, Alibaba, Tencent, and NetEase Cloud Music to withdraw or limit companion features. The measures illustrate China’s preference for targeted, service-specific controls while a broader AI law remains under discussion.
The updated account sharpens the rules’ substantive prohibitions, including manipulation, self-harm encouragement, and certain virtual intimacy, while placing them within a continuing debate over broader Chinese AI legislation. It also identifies the State Council as a relevant actor and adds a reported three-month compliance period.
The main update is clearer evidence of implementation effects: major platforms reportedly rolled back or shut down companion features as the rules took effect. The framing also tightens around disclosure, minors, and product designs built on persistent memory and continuous emotional engagement.
The story has shifted from a broad description of China’s emerging companion-AI rules to confirmation that the measures are now in force and are being actively enforced. The current version also adds a broader comparative frame, contrasting China’s approval-and-filing approach with U.S. and EU AI oversight models.
- Rules now explicitly in force since July 15, 2026.
- Safety reviews required after major changes and user-threshold triggers.
- Sensitive companion-interaction data banned from training without consent.
- Shanghai removed more than 14,000 non-compliant agents.
- U.S. and EU AI oversight models added for comparison.
China’s AI story has shifted from a broad move toward unified regulation to an immediate enforcement phase for anthropomorphic companion services. The new rules create concrete operating constraints and have already triggered reported product rollbacks, while the planned unified AI law remains the longer-term consolidation effort.
- Anthropomorphic AI rules took effect on July 15, 2026.
- The rules target sustained emotional interaction services, not ordinary assistants.
- Users under 18 are barred from virtual intimate relationship services.
- App stores must verify filings and can suspend or remove noncompliant services.
- ByteDance, Alibaba, Tencent, and NetEase reportedly rolled back companion features.
The story now places greater emphasis on China moving toward a first comprehensive AI law, rather than just a broader national framework. The policy-analysis angle is also slightly reframed as evidence of a still-uneven governance system that Beijing is tightening under state-control and U.S.-China competition pressures.
The material centers on how China is structuring AI governance through central-level policy design, quantitative policy evaluation, and movement toward a unified AI law. Across the texts, the focus is less on individual AI applications than on the coherence, coverage, and institutional design of the regulatory system. The dominant finding is that China’s framework is becoming more systematized, while gaps remain in long-term planning, coordination, and policy instrument balance.