Platforms would bear primary responsibility, deploying classifiers to detect cyberviolence, offering users one-click protection, restricting abusive content, and investigating offending accounts.

Tech platforms and AI service providers would face accountability measures, including classification rules for online environments and potential criminal liability for non-compliance.

Zhihu has begun curbing anonymous accounts in anticipation of the rules, signaling rapid platform compliance even before finalization.

The draft aims to protect victims while increasing state oversight, raising privacy and surveillance concerns due to required logging, tracing, and blacklist systems.

Enforcement includes offender blacklists, regular intervention reports, early-warning systems to curb pile-ons, and penalties up to account closures for abusers.

China’s cyberspace regulator has released a draft law to curb cyberbullying, including AI-enabled abuse, applying to activities within China and to overseas actors targeting the country.

The draft targets both AI-generated content and broader online safety responsibilities, reflecting Beijing’s tightening approach to online governance.

Online platforms would be required to identify and remove or block AI-generated abusive content, report it to authorities, and extend rules to foreign entities whose actions target Beijing.

Public comments are open, and the final text may evolve, but the direction is clear: platforms must monitor and be accountable for online abuse at scale, including AI-driven content.

The draft targets malicious actors using automated tools, troll armies, or coordinated AI-driven harassment campaigns as apparent criminal liability.

Platforms must verify user identities before posting or using messaging services, trace or report AI-generated risk content, and remove or block AI-driven cyberbullying content.

A major provision requires real-name verification for users before posting or accessing messaging services.