China’s cyberbullying draft makes AI-enabled abuse a legal category

What victim-centered enforcement requires

AI changes the scale and persistence of abuse. A single perpetrator can generate many convincing artifacts, imitate a target, automate distribution, and re-create removed material. Victims therefore need fast evidence preservation, clear reporting routes, meaningful response times, limits on re-uploading, and remedies that do not require them to repeatedly prove the same harm.

Speed must be paired with due process. Legislators and platforms should distinguish targeted abuse from criticism, satire, documentation, and public-interest reporting; disclose when automated systems make enforcement decisions; preserve appeal rights; and publish enough aggregate information for the public to see whether the law protects people consistently.