Cyberviolence Law overview

1. Background

China has gradually been expanding the legal system’s capacity to address Cyberviolence through a series of normative documents and has now released a draft of its first Cyberviolence Law.

Notice on Effectively Strengthening the Governance of Cyberviolence ("Notice")

The CAC's 2022 Notice on Effectively Strengthening the Governance of Cyberviolence was the first authority specifically dedicated to cyberviolence. It spelled out online platforms' obligations in preventing and addressing cyberviolence in substantial detail.

Guiding Opinions on Punishing Cyberviolence Violations and Crimes in Accordance with Law ("Guiding Opinions")

The leading justice sector authorities, the Supreme People's Court, Supreme People's Procuratorate, and Ministry of Public Security, next released these detailed opinions describing how certain criminal charges and administrative violations should be applied in the context of cyberviolence.

Provisions on the Governance of Cyberviolence Information ("Provisions")

Jointly released by the CAC, MPS, Ministry of Culture and Tourism, and State Administration of Radio and Television; these Provisions attempt to create a more comprehensive system addressing for addressing and preventing Cyberviolence- going beyond online platform's responsibilities and justice sector penalties to include a broader societal approach.

Law on Countering Cyberviolence （Draft）

Now, a draft Cyberviolence Law is available for public comment until 8/28/2026. It integrates the previous documents, most directly building off the broader mandate of the Provisions, and adding in departments for education, civil affairs, and telecommunications work in its prevention and governance work.
It does not completely supersede earlier authority, however, and at times goes into less granular detail than the more focused Notice and Guiding Opinions, such as on how specific charges apply.

A table at the end of this post includes my notes comparing prior authority with the draft.

The draft Law is largely organized around the obligations and roles of different parties.

• Chapter 2 addresses the obligations of online service providers (platforms).
• Chapter 3 focuses on the role of government agencies,
• Chapter 4 is on societal governance, addressing the responsibilities of industry associations, civil society groups, internet users, and other individuals and organizations.
• Chapter 5 switches to the justic sector: the courts, procuratorate, and police. 

This structure, similar to that of the Law on the Protection of Minors is fairly intuitive, but is not strictly followed, leading to some odd outcomes, like scattering the articulation of offenses throughout the law.（See for example, articles 4, 31, 32, etc)

2. Definition of Cyberviolence

The definition of cyberviolence has largely remained consistent as the legal authority has evolved, emphasizing injury to individuals’ privacy and personality rights (such as reputation, honor, likeness） as well as disruptions of order online.

Practically, this has included a wide range of conduct, such as doxing, spreading rumors, revenge porn, pile-on attacks, hiring professional troll armies, etc.

The draft law follows the early Provisions in including language that cyberviolence should be serious, by saying that it should be ‘concentrated’ or ‘continuous’, emphasizing scale or duration, and lists four categories in its definition at article 2.

1.  Concentrated publishing of “Cyberviolence Information,”  which includes insults, rumor-mongering, defamation, incitements of hatred, encouraging antagonism, threats and coercion, discrimination, or prejudice, etc. This largely follows the definition in Article 32 of the Provisions, but some categories such as ‘encouraging antagonism’ and inciting prejudice were adapted from internet content regulations and not previously labeled as cyberviolence.
2. Illegally publishing others’ personal information in a concentrated manner. Building on the Personal Information Protection Law, with specific offenses detailed already in the Guiding Opinions on Addressing Violations.
3. Intimidation and Harassment. This was included in the Provisions’ definition of Cyberviolence Information, and it is not clear why they were broken out here, or if doing so makes sense. It could be an attempt to distinguish direct attacks from the publication of harmful information, but the line is hazy as online life is conducted entirely through information exchanges.
4. Other activities that harm rights and interests. Catch-all clauses like this are common in Chinese law, but the vagueness can be problematic. Here, we know that the covered conduct must be tortious.

2.1 Prohibited Conduct

In addition to the definition, a few lists of prohibited conduct also help explain the scope of ‘cyberviolence’

Article 31, for example, contains activities that are forbidden when used for cyberviolence, seemingly focusing on ways of ramping up the scale of attacks:

1. Registering fake or batch accounts, stirring up viral public opinion topics, and manipulating traffic.
2. Manipulating traffic through fake clicks, votes, lists, rankings, comments, or ratings; traffic hijacking, heating or priming content to get more views..
3. Illegally handling personal information that was legally disclosed. This is new to the law, allowing that even where personal information has become publicly available via consent, self-publication, etc., some types of handling are still prohibited. That might include use beyond the scope for which information was released, use in rumors or defamation, etc.
4. Using AI  to carry out cyberviolence (generation, synthesis, or personalized recommendations); Early cyberviolence authority considered the use of AI an aggravating factor in some crimes or an enforcement priority, but listing it as a separate prohibition is new. In addition to using AI for content generation, AI has also been used to expand cyberviolence attacks with bot armies. SEE e.g.

Article 32 more closely follows existing law to prohibit knowingly assisting the perpetrators of Cyberviolence. This ‘assistance’ can be in the form of

1. Technical support: internet access, hosting, cloud storage, data transmission.
2. Financial support: through means such as monetized traffic, advertising, promotion, or payment processing. [similar to rules for supporting other illegal content providers- like obscenity.
3. Providing Personal Information: there are already offenses for illegally sharing/selling personal information, but doing so while knowing it will be used for cyberviolence adds another level of culpability.
4.  Support in evading real name systems: By providing phone cards, IoT cards, Internet accounts, etc.

3. Other Notable New Content.

1.  Enforcement Tools 
   1. Written warnings: [告诫书] have long been used to spur compliance in the regulatory context, but were introduced into individual conflicts in the 2015 Domestic Violence Law. The idea is that they are both a deterrent and create an evidentiary record for future enforcement if the problem continues. [article 40]
   2. Injunctions: The Guiding Opinions mentioned the use of injunctions in article 15, but this is expanded on significantly in the draft’s article 37. Parties are able to go directly to the court for a protective injunction when they can demonstrate cyberviolence or an imminent threat of cyberviolence that could create a difficult-to-remedy harm. Where the victim is a minor or lacks civil capacity, mass organizations, like the women’s federations or villagers and resident committees, may apply for injunctions on their behalf. The injunctions are enforced by the courts, with the public security organs and other departments assisting.
2. Protections for vulnerable populations- The draft law is cognizant of vulnerable groups throughout, and provides special protections/rules for them, particularly minors.
   1. Priority: While the Provisions mentioned that platforms should prioritize minors’ reports, Article 36 of the draft law expands this to also include other vulnerable groups, such as the elderly. Platforms are instructed to generally be more proactive in protecting these groups. (14)
   2. Procuratorate support in private prosecutions: Where minors or persons with disabilities are launching a private prosecution, meaning that the procuratorate is not prosecuting in the name of the state, they may still assist these groups in pursuing the charges. (41)
   3. Guardian notifications: Minors’ guardians should be able to receive cyberviolence risk alerts and education. (14)
3. Mental/Emotional Damages compensation: now expressly allowed in civil litigation regarding cyberviolence. (57) In addition, the state is to support the provision of social workers to assist victims (35).
4. MCN- Multi-Channel Networks: refers to cross-platform content distribution services that ‘sign’ online talent and help with marketing and promotions. They have been showing up increasingly in Chinese regulations, and have special duties in the draft to discipline the accounts they manage that are involved in cyberviolence. (34) MCNs can face blacklisting where they are directly involved in Cyberviolence or encourage it.

4. Table

Jeremy Daum

• Jeremy Daum

  

  2026/09/21

  Quicktake: Draft Rules for Protection of Minors Online
• Jeremy Daum

  

  2026/08/05

  Brief Notes on Public Interest Litigation Law (and comparison of Drafts 1/2)
• Jeremy Daum

  

  2026/01/20

  Alien Chat: China’s 1st Case of Criminal Responsibility for AI-Generated Content
• Jeremy Daum

  

  2026/01/12

  Did China make sexting illegal?