China proposes cyberbullying law covering AI-generated abuse

China’s cyberspace regulator released a draft cyberbullying law on July 29 that would explicitly cover abuse created or spread with artificial intelligence, including synthetic content designed to humiliate or intimidate people at scale. The proposal comes as generative AI has made deepfakes, impersonation and coordinated harassment campaigns easier to produce and harder to contain.

The draft would draw a clearer legal line around cyberbullying and push platforms and technology companies to take more responsibility for what circulates on their services. Follow-on summaries of the proposal said online platforms would have to verify users’ real identities before allowing them to post information. Another summary said platforms would be required to monitor and remove harmful synthetic content or face fines.

That enforcement model fits Beijing’s broader campaign to police online discourse, protect minors and limit harmful content. It also extends a regulatory trajectory that has been building for years. China had already issued draft guidelines to tackle online bullying on June 9, 2023, and the country’s Personal Information Protection Law took effect on Nov. 1, 2021. Together with rules on generative AI and deep synthesis, the new draft shows regulators folding synthetic content, impersonation and algorithmic abuse into the same speech-control framework.

The practical effect would fall most heavily on Chinese social media companies, AI developers and other user-generated platforms. Real-name verification would make anonymous posting harder in some contexts, while monitoring and removal rules would raise the cost of hosting edgy or borderline content that could be weaponized into harassment. For users targeted by doctored images, fake voices or AI-written smear campaigns, the draft points to a system that treats machine-made abuse as a compliance issue, not just a moderation problem.

The proposal also highlights how sharply China’s approach differs from the slower, more fragmented response in the U.S. and Europe, where AI-harm debates have moved through separate privacy, safety and platform enforcement fights rather than a single cyberbullying law. In Beijing’s framework, the responsibility does not stop at the person who posts the abuse. It reaches the platforms and the tools that help generate it, which is where regulators appear determined to set the rules before the damage scales further.