Overview of New Regulations
China’s Cyberspace Administration (CAC) has proposed two new draft regulations aimed at managing AI-generated content. These drafts focus on labelling AI content to combat deepfake risks and ensure that information remains credible. The regulations primarily target internet service providers, including those generating and disseminating AI content. Notably, they also extend to foreign service providers targeting users in Mainland China.
Key Details of the Draft Regulations
- The Draft AI Labelling Measures mandate both explicit and implicit labelling of AI-generated content. Explicit labels are visible to users, while implicit labels are hidden in metadata.
- Service providers must integrate labelling into user agreements and submit related materials during security assessments.
- Different labelling actions are required based on the certainty of whether content is AI-generated.
- App stores and users also share labelling responsibilities, with app stores needing to verify compliance before app publication.
- Although specific penalties for violations are not detailed, enforcement will be managed by the CAC, which can impose penalties based on existing laws.
Importance of These Regulations
These regulations are crucial for enhancing transparency and accountability in the AI sector. By requiring clear labelling, they aim to reduce the spread of misinformation and protect the rights of content creators and consumers. As global standards for AI governance evolve, these measures position China as a leader in AI regulation. They also create challenges for international players who must navigate various compliance requirements across different jurisdictions. The implications of these regulations will likely influence the development of AI detection technologies and shape the future landscape of AI content management worldwide.
