China's CAC Completes Phase 1 of 'Qinglang·AI Application Chaos' Crackdown, Targets Fake News and Minors' Protection in Phase 2

On July 6, the Cyberspace Administration of China (CAC) announced the completion of Phase 1 of its "Qinglang·Rectifying AI Application Chaos" special campaign, marking a significant escalation in China's AI governance framework.

While specific results from Phase 1 were not detailed in the announcement, the CAC laid out an aggressive agenda for Phase 2, targeting six major problem areas in AI application misuse.

The priority targets include: using AI technology to produce and distribute false information, spreading violent and vulgar content, identity impersonation and deepfake fraud, infringing on the legal rights of minors, and operating AI-powered "water army" (paid poster) networks.

On enforcement, the CAC pledged to increase penalties against violators — both individual accounts and institutional actors — creating meaningful deterrence. The regulator also plans to strengthen oversight across key AI application touchpoints in the content production and distribution chain.

Platform responsibility is a central pillar of the new phase. The CAC explicitly required internet platforms to improve their AI risk prevention capabilities, establishing more robust content review and risk control mechanisms to curb AI abuse at both the technical and managerial levels.

China has been steadily tightening its AI regulatory framework since 2025. The "Qinglang" campaign series, which previously targeted everything from celebrity gossip to financial misinformation, now extends its reach into the AI application layer, signaling a shift from principle-based guidelines to targeted enforcement operations.

As AI-generated content becomes increasingly embedded in social media, news dissemination, and commercial marketing, regulators worldwide are grappling with the same challenge: how to balance technological innovation with safeguards against misuse. China's approach in this campaign could offer a template — or a cautionary tale — for other jurisdictions pursuing AI governance.