Internet & Online Safety ยท China
Online safety & content laws in China (2026)
China shaded by its internet & online safety status
Online safety rules in China: heavy restriction.
FrameworkCybersecurity Law (2017; amended effective 1 Jan 2026) plus a dense body of CAC regulations โ including the Administrative Measures for Internet Information Services, Regulations on Ecological Governance of Network Information Content, Algorithmic Recommendation Provisions, Deep Synthesis Provisions, AI-Generated Content Labeling Measures, and Regulations on the Protection of Minors Online โ enforced by the Cyberspace Administration of China (CAC) alongside MIIT, MPS and NRTA, and reinforced by the 'Great Firewall' technical censorship system.
China operates the world's most extensive state-run online-content control regime. The Cyberspace Administration of China (CAC) sits at the center of a layered legal framework โ the Cybersecurity Law (whose most significant amendment since 2017 takes effect 1 January 2026), sector-specific CAC measures on algorithms, deep synthesis, AI-generated content and minors, and a technical censorship apparatus (the 'Great Firewall') that blocks most major foreign platforms (Google, Facebook, YouTube, X, most Western news sites) and criminalizes unauthorized VPN use. Platforms bear strict primary liability to actively police, filter, real-name-verify users and remove prohibited content.
Key points
The Cybersecurity Law (in force since June 2017) is the anchor statute; on 28 October 2025 the NPC Standing Committee adopted the biggest amendment since its adoption, taking effect 1 January 2026. It designates the CAC as the coordinating authority, sharply increases penalties, and adds dedicated AI-governance provisions. Articles 47โ52 require network operators to monitor, block and report prohibited user content and to comply with government take-down and blocking orders.
China maintains the world's most sophisticated national content filtering system, using DNS poisoning, IP blocking, deep packet inspection and TCP-reset injection to block Google, Facebook, YouTube, X/Twitter, Instagram, WhatsApp, BBC, Reuters, CNN and thousands of other services. Unauthorized VPN use to circumvent the block is illegal; enforcement tightened further in 2026, and a August 2025 GFW incident briefly disrupted essentially all outbound HTTPS traffic.
Platforms have primary liability: they must implement keyword filtering, real-time monitoring, pre-publication review of generated content, maintain a CAC-updated prohibited-content database, and immediately halt transmission of illegal content and report it. New Regulations on Internet Content Multi-Channel Network Distribution Services enter into force 1 September 2026, and CAC issued a further revised draft of the Administrative Measures for Internet Information Services on 3 July 2026 extending platform accountability and executive personal-liability mechanisms.
Real-name registration has been mandatory for internet, messaging and social services since 2017 and was extended in 2025 with a national online identity authentication system issuing 'Internet certificates' as a centrally verified credential. Games, live streaming, video and social platforms must verify every user's real identity; anonymous posting on public-facing information services is prohibited.
The Regulations on the Protection of Minors Online (issued 24 Oct 2023, effective 1 Jan 2024) impose strict duties on operators serving users under 18: mandatory guardian-verified real-name checks, curfews and time limits on gaming and live streaming, anti-addiction systems, and prohibitions on renting/selling minor gaming accounts. On 16 Sept 2025 CAC issued draft Measures identifying 'platform service providers with a substantial number of minor users and significant influence' for heightened obligations.
China regulates recommendation algorithms (Algorithmic Recommendation Provisions, 2022), deepfakes/synthetic media (Deep Synthesis Provisions, 2023) and generative AI (Interim Measures for Generative AI Services, 2023). The joint CAC/MIIT/MPS/NRTA Measures for Labeling AI-Generated and Synthetic Content, issued 14 March 2025 and effective 1 September 2025, require both explicit (user-visible) and implicit (metadata) labels on all AI-generated text, image, audio, video and virtual-scene content.
Timeline - major decisions & events
The first revision since 2017 (adopted by the NPC Standing Committee on 2025-10-28) raises maximum fines to RMB 10 million, expands extraterritorial reach to any activity endangering China's cybersecurity, and adds an article on safe AI development. It signals tighter integration of content, data, and AI governance.
State Council Information Office โCAC's Measures for Labeling AI-Generated Synthetic Content (released 2025-03-14) and national standard GB 45438-2025 require both explicit on-screen labels and implicit watermarks/metadata on AI text, images, audio, and video distributed on Chinese platforms. It is one of the world's first comprehensive AI-labeling mandates.
China Law Translate (CAC measures) โCAC issued guidelines for a unified 'minor mode' across devices, app stores, and apps, automatically filtering content and limiting screen time by age. It operationalizes the 2024 minors regulation across the mobile ecosystem.
China Law Translate (CAC guidelines) โThe State Council regulation (released 2023-10-16), comprising 60 articles, requires platforms to curb addiction, restrict harmful content, and protect minors' personal data, with CAC coordinating enforcement. It is China's first dedicated administrative regulation on children's online safety.
Gov.cn (State Council) โIssued 2023-07-10 by CAC and six other departments, these rules require generative AI providers to ensure content aligns with 'core socialist values,' label synthetic output, and conduct security assessments. They were the first national rules specifically governing generative AI services.
Library of Congress โReleased by CAC and others on 2021-12-31, the rules require recommendation-algorithm providers to register algorithms, offer opt-outs, avoid price discrimination, and promote 'positive energy' content. They marked China's move to directly regulate platform algorithms shaping what users see.
Bird & Bird (CAC provisions) โAdopted 2021-08-20, the PIPL is China's first comprehensive data-privacy law, governing consent, cross-border transfers, and platform handling of personal information, with GDPR-like extraterritorial reach. It completes the CSL-DSL-PIPL data governance triad underpinning online regulation.
National People's Congress โAdopted by the NPC on 2021-06-10, the DSL establishes a national data-classification and security framework treating data as a national-security concern and restricting cross-border data transfers. It expanded state control over data flowing through online services.
National People's Congress โAdopted 2016-11-07, the CSL is the foundational statute for China's internet governance, mandating real-name registration, data localization for critical infrastructure, and platform responsibility to police prohibited content. It centralized CAC's authority over the online sphere.
DigiChina (Stanford) โThe NPC Standing Committee required network service providers to collect users' real identities and to delete, record, and report prohibited content to authorities. It cemented real-name registration and intermediary-liability principles later codified in the Cybersecurity Law.
Library of Congress โState Council Order No. 292 created the licensing/filing regime for internet information services, dividing them into commercial and non-commercial categories and prohibiting publication of broad categories of content. It established the basic legal scaffolding for content regulation online.
DigiChina (Stanford) โThe Ministry of Public Security began building a national content-filtering and surveillance system that blocks foreign sites and politically sensitive material at China's internet gateways. It remains the technical backbone of China's online content control.
Encyclopaedia Britannica โChina - other topics
Internet & Online Safety in other countries
Last verified 8/27/2026 ยท Orientation, not legal advice - verify against the primary sources linked above. Methodology & how to cite ยท State of Technology Regulation 2026 ยท Explore the full world map โ