What happened
On April 10, 2026, China's Cyberspace Administration (CAC), together with four co-issuing ministries (NDRC, MIIT, MPS, SAMR), published the Interim Measures for the Administration of Artificial Intelligence Anthropomorphic Interaction Services, effective July 15, 2026. The measures cover virtual companions, emotionally responsive chatbots, AI-powered digital humans, and any service that simulates human-like interaction.
Why it matters
This is China's most comprehensive regulation targeting AI companionship and emotionally interactive AI — and it carries extraterritorial reach for any offshore provider with Chinese users. Key obligations include mandatory AI-disclosure pop-ups, a two-hour usage break prompt, a full prohibition on virtual companion or 'virtual relative' services to under-18s, algorithm filing, security assessment, and CAC registration once user thresholds are crossed (1 million registered or 100,000 monthly active users).
Action needed
Organisations with consumer-facing AI products serving mainland Chinese users should conduct a gap analysis against the July 15 compliance deadline. Priority checks: content labelling, minor-user safeguards, and whether your product crosses the registration thresholds that trigger mandatory security assessment submission.