China Proposes Anti-Cyberbullying Law Targeting AI-Generated Abuse, Signaling Broader Push to Regulate AI Risks

China is moving to tighten oversight of artificial intelligence and online platforms with draft legislation aimed at combating cyberbullying, including abuse generated or amplified by AI, in the latest sign that Beijing is taking an increasingly proactive approach to regulating the rapidly evolving technology.

The draft law, released Wednesday by the Cyberspace Administration of China (CAC), would impose new obligations on internet platforms to detect, remove and report AI-enabled cyberbullying, while expanding user verification requirements and introducing tougher penalties for companies that fail to comply.

The proposal comes as governments around the world grapple with the societal risks posed by increasingly powerful AI systems. While the United States has largely relied on a mix of executive actions, agency guidance and sector-specific initiatives, China has moved more aggressively to establish binding rules governing AI technologies before they become deeply embedded across society.

Beijing has already introduced regulations covering generative AI services, recommendation algorithms and deep synthesis technologies. The latest proposal broadens that regulatory framework by directly addressing how AI can be used to facilitate online harassment and abuse.

If enacted, the law would apply not only to cyberbullying activities carried out within China but also to organizations and individuals overseas whose online activities target the country.

One of the proposal's most notable provisions is its explicit focus on AI-enabled cyberbullying.

Online platforms would be required to identify, trace and assess cyberbullying content generated or disseminated through artificial intelligence. Companies would also be obligated to promptly remove or block such content and to report serious cases to government authorities.

The measure emerges from growing global concern that generative AI can dramatically increase the scale and sophistication of online abuse through fake images, manipulated videos, cloned voices, automated harassment campaigns and other forms of synthetic media.

Unlike traditional moderation rules, the draft specifically recognizes AI as a force multiplier for harmful online behavior, suggesting regulators are attempting to anticipate emerging risks rather than responding only after they become widespread.

The proposal also bolsters China's longstanding emphasis on platform accountability. Internet companies would be required to verify users' real identities before allowing them to publish content or use instant messaging services, further strengthening China's real-name registration system.

By tying online activity to verified identities, authorities aim to make anonymous harassment more difficult while increasing accountability for abusive behavior.

Schools would also be required to incorporate anti-cyberbullying education into their curricula, signaling that Beijing views online safety as both a regulatory and social issue requiring preventive education alongside enforcement.

The draft law provides regulators with broad enforcement powers. Online service providers that violate the rules could face fines of up to 10 million yuan (about $1.5 million). Authorities would also have the power to suspend websites or applications or revoke business licenses in serious cases.

The Cyberspace Administration of China is accepting public comments on the proposal until August 28 before moving toward final implementation.

China Takes A More Proactive Regulatory Path

The proposal points to a broader divergence between how China and the United States are approaching AI governance.

China has consistently sought to regulate AI technologies early through comprehensive national rules that place clear legal obligations on developers and platforms. Its regulatory plan has focused on establishing guardrails before technologies reach mass adoption, particularly in areas involving content generation, online safety, data governance and national security.

By contrast, the United States has generally taken a more decentralized and market-oriented approach. Federal oversight has largely been shaped by executive orders, voluntary commitments from AI companies, agency-specific guidance and existing laws governing privacy, competition and consumer protection. While lawmakers have introduced numerous AI-related bills, Congress has yet to enact a comprehensive federal AI law, leaving regulation fragmented across agencies and states.

The latest Chinese proposal therefore reinforces Beijing's position as one of the world's most active AI regulators. Rather than waiting for harms to become widespread, authorities are increasingly attempting to codify rules governing how AI systems are developed, deployed and monitored.

The focus on AI-enabled cyberbullying also underpins a shift in AI policymaking globally. Early debates centered largely on issues such as copyright, misinformation and employment. Regulators are now expanding their attention to the social consequences of capable AI systems, including their potential to automate harassment, impersonation and coordinated abuse at a scale previously impossible.