New Legislation in China Regarding Generative AI Restructuring

China has recently introduced a law This law aims to reshape the internet landscape in China through the use of generative AI technologies. It's set to launch by the year's end and specifies the rules for using generative AI, backed by severe repercussions for noncompliance. The legislation highlights which uses of generative AI are deemed unlawful, leading to notable repercussions.

The legislation categorizes any use of generative AI as illegal if its training data exceeds 5% of content deemed \"illegal and harmful.\" In the context of China's regulations, “harmful content” this refers to information that is restricted by the Great Firewall of China.

Consequently, beginning next year, a major transformation of China's online environment will start, gradually purging 'harmful content.' This shift will be led by the increased deployment of generative AI, where developers of these AI frameworks will act as the architects of this digital evolution. The state will rigorously oversee this reformatting initiative’s results.

This extensive overhaul encompasses historical, contemporary, and predictive data, alongside journalism and literature. The restructuring will largely depend on advanced formatting tools, primarily powered by generative AI, with developers evolving into specialists in formatting practices. The government will impose strict controls over the results of this considerable reformatting project, marking a significant change in content regulation in China.

Given the rapid proliferation of generative AI tools in numerous platforms, it is expected that within just a few years, the digital landscape of China will be thoroughly transformed, with certain elements left in the past.

• In August, China has established detailed guidelines for the deployment of generative AI applications that require AI service providers to maintain transparency in accountability, moderation of content, data handling, and safeguarding user data. These regulations aim to foster the growth of the AI sector while ensuring adherence to compliance.
• Furthermore, China’s Cyberspace Administration has declared that stringent licensing protocols will be imposed on generative AI technologies which are seen as capable of supplanting human jobs. The Administration has published a compulsory list of 41 generative AI algorithms that must be licensed prior to their general application, impacting all 79 major Chinese generative AI language models. LLMs ) developed based on these algorithms.