China mandates AI content labels. Platforms test label visibility

China's rules on labeling AI-generated content took effect on September 1, 2025. The regulation requires two types of labels: an explicit one visible to the user, and a hidden one in file metadata identifying the model provider and generation time.

Sina checked how a dozen platforms implement the rules. Results vary. Douyin, ByteDance's video app, added a checkbox at publishing and a watermark for content marked as generated. WeChat Channels let authors declare AI use. If the platform suspects AI content, it triggers a manual review. Weibo placed the label in the share menu, with the wording "content may be generated by AI." Toutiao introduced automatic labeling and stores metadata with the provider code and creation time.

Not every service made the deadline. Several smaller apps had no mechanism in place when Sina ran its test. The government rolled out a national tool for verifying labels, along with technical standard GB 45438-2025, which defines metadata storage. The maximum penalty for a missing label is 100,000 yuan. For large platforms, that is a symbolic sum. Reputational risk and the prospect of throttled reach already hurt.

Reach as a sanction

Throttling reach proved more effective than the fine. On August 29, Douyin announced that content without the required label would get reduced visibility. Repeated violations could lead to an account ban. This changed the cost calculation. For a creator whose income depends on views, losing distribution is a far harsher punishment than any fine.

Tencent's double standard

Outlet 36Kr described a separate case. Tencent is testing an AI-based writing assistant in its Moments feature (朋友圈). At the same time, it is tightening rules for AI content in official public accounts. Channels run by companies and creators must declare model use. Posts not generated by a human have their reach limited. Newsrooms in China consider this the most restrictive approach on the market. It is not a ban on using AI, but an obligation to separate private conversation from publication.

Financial context shows why Tencent can afford caution. In the second quarter of 2026, the company's capital expenditure rose 176 percent year on year, to 52.78 billion yuan. Free cash flow was negative at 13.8 billion yuan. At that scale of investment in AI infrastructure, regulatory risk becomes a strategic variable, not an operational one.

A lesson for European newsrooms

The Chinese model shows that a label alone is not enough. What matters is who sees it and how quickly. A label hidden in the share menu meets the formal requirement but does not help the audience distinguish real material from generated material. Effectiveness comes from combining three elements: automatic labeling in metadata, a visible mark in the interface, and a real sanction for omission. The EU AI Act moves in a similar direction, but provides for longer transition periods and milder penalties for platforms.