On 17 November 2023 the Central Cyberspace Affairs Commission issued a notice launching a one-month nationwide campaign against online hostility. The campaign focuses on social, short-video and live-streaming platforms, targeting seven categories of conduct including anonymous shaming pages, doxxing, malicious attacks and rumour-mongering around news events, stigmatising groups and inciting regional conflict.
Such campaigns do not create new legal obligations directly, but they change enforcement intensity. During a campaign, platforms face stricter expectations on response speed, severity of action and evidence.
Why “hostility”
Hostility is not a single unlawful act but a pattern: attacks, humiliation, pile-ons, exposure of private information and incitement of conflict. Individually they may look like merely inappropriate speech; combined, they cause sustained harm and can spill over into offline risk.
The difficulty for platforms is the boundary. Criticism and attack, banter and humiliation, public debate and group conflict are not separated by an automatically detectable line. The campaign requires platforms to write that line into their rules and apply it in a reviewable way.
What the seven categories mean
First, anonymous shaming pages and doxxing. These accounts aim to expose and humiliate specific people, often with personal information. Platforms need targeted detection rules and fast channels, linked to personal-information protection mechanisms.
Second, malicious attacks and rumours around news events. Traffic concentrates during breaking events and rumours spread quickly. Platforms need temporary review strategies and fact-checking mechanisms for such periods.
Third, stigmatising groups and inciting regional conflict. Such content often appears as memes and is hard to detect. Platforms need semantic detection plus human review, not keywords alone.
The remaining categories involve extreme emotional venting, organised pile-ons and inducing minors to participate. The common feature: a single post may not breach rules, but account behaviour is persistent.
Three adjustments for platforms
First, move from content enforcement to account enforcement. Deleting individual posts has limited effect against persistent hostility. Platforms need account-level scoring, throttling, suspension and bans, with records of action.
Second, move from after-the-fact deletion to scenario prevention. For news events, live interaction and comment sections, pre-configure review strategies and human coverage.
Third, move from going it alone to coordinated handling. Where personal information is leaked, offline threats arise or organised pile-ons occur, platforms need notification and cooperation channels with cyberspace and public security authorities.
Rules and evidence must move together
The most common mistake during a campaign is taking action without being able to explain it. Inspectors look not only at outcomes but at basis and process.
Platforms should ensure that community rules clearly describe the seven categories; that enforcement maps to specific rule clauses; that appeals work; that records are traceable; and that errors can be corrected. Without these, tougher enforcement produces more disputes.
Minors are a priority scenario
Hostile content affects minors directly. Platforms should check whether minors’ mode genuinely isolates high-risk content, whether private messages, groups and comment sections contain entry points that induce participation in pile-ons, and whether reporting channels are minor-friendly.
For platform governance and data compliance, see our data compliance service; for community rules, user terms and dispute handling, see commercial contracts and commercial dispute; for ongoing support, see annual legal counsel.
This article is based on the notice on the Qinglang campaign against online hostility issued by the Central Cyberspace Affairs Commission on 17 November 2023. It is for reference only and does not constitute legal advice.