AI Regulation Tracker / Draft consultation
China proposes a dedicated AI chapter in its revised internet information rules
On July 3, 2026, China's Cyberspace Administration published a revised draft of the Measures for the Administration of Internet Information Services (互联网信息服务管理办法) for a second round of public comment. The draft would add a bespoke Intelligent Information Services section. This is a proposal, open for comment until August 2.
China already regulates internet information services and, separately, algorithms and deep synthesis. What this draft proposes is to pull AI-specific duties into the core internet-information-service rules for the first time, through a section written for the purpose. On July 3, 2026, the CAC published the revised draft and reopened public comment, the second such round for this rewrite. The instrument is the Measures for the Administration of Internet Information Services, and the headline change for AI work is a proposed section on Intelligent Information Services.
What the draft would add
The draft would create a dedicated section addressing services that apply AI technology to provide internet information. The state notice describes it this way: 增设"智能信息服务"专节，对应用人工智能技术提供互联网信息服务的情形进行专门规范. In plain English, that adds an Intelligent Information Services section and sets out rules specifically for the situation where AI is used to deliver internet information services. If finalized, it would be the first bespoke AI regime placed inside the internet-information-service rules themselves rather than in a standalone algorithm or deep-synthesis measure.
Alongside the new section, state and secondary reporting describe several sub-duties the draft would carry. As reported, these include labeling of AI-generated content, human-intervention mechanisms for algorithmic recommendation, tightened account management, and a serious-dishonesty list. The draft notice sets out the account-management and dishonesty pieces directly: 建立互联网领域严重失信主体名单管理制度，增设完善限制账号功能、禁止注册新用户账号等管理措施. Treat the itemized AI sub-duties as reported rather than line-quoted, because the operative wording will only be fixed if and when the measure is finalized.
A draft for comment, not a rule in force
This is the distinction that matters for planning. Publication on July 3 opens a consultation. It does not create any obligation. The CAC has set a comment deadline of August 2, 2026, with feedback directed to law@cac.gov.cn, and this is the second round, which means the text has already moved once and can move again. Anyone describing companies as now required to label AI content or to maintain a dishonesty list under this instrument is describing a proposal that has not been adopted. The correct read today is that the CAC would require these things only if the draft is finalized in something like its current form.
Why a US professional should track this
The direct audience is any organization that provides internet information services in China, which includes US technology and AI companies with China-facing products, local entities, or platform partnerships. If that is you, the comment window is the moment to read the draft against your own services and decide whether to file. The Intelligent Information Services section is the part to map first, because it would define which of your AI-driven features fall inside the regime.
There is a second reason to watch it. China is proposing to fold AI-specific duties directly into its baseline internet rules rather than leaving them in separate algorithm measures. That structural choice is itself a signal. Other jurisdictions weighing how to house AI obligations, inside existing platform law or in standalone AI statutes, will study how China draws this section. That makes the draft an early data point on regulatory architecture, not only a domestic Chinese proposal.
Questions professionals are asking
Does China now require AI-content labeling under these Measures?
No. This is a revised draft published for public comment on July 3, 2026, with comment open until August 2. Reported duties such as AI-content labeling would apply only if the draft is finalized, and the wording could change before then.
What is the Intelligent Information Services section?
It is a proposed new section (智能信息服务专节) that would set rules specifically for services using AI technology to provide internet information. If finalized, it would be the first bespoke AI regime inside China's internet-information-service rules rather than in a separate algorithm measure.
Does this affect a US company?
It would affect US organizations that provide internet information services in China, through local products, entities, or platform partnerships. The practical step now is to read the draft, map which AI-driven features would fall inside the new section, and decide whether to file comment before August 2.
RELATED BRIEFINGS
Browse the full AI Regulation News tracker
Informational analysis for working professionals, not legal advice. Confirm how any rule applies to your situation with qualified counsel in the relevant jurisdiction.