China has recently issued new regulations on generative artificial intelligence (AI) models, taking a lead in establishing comprehensive guidelines for AI tools such as OpenAI’s ChatGPT. Coming into effect on August 15, these rules will apply to all generative AI content services in China, including text, images, audio, and video. The move has sparked considerable discussion about the future of AI in the country.
Table of Contents
AI Policy of China : Regulating Generative AI
AI policy, enacted by the Cyberspace Administration of China and jointly published by seven Chinese regulators, is the first of its kind. The law has a pronounced focus on promoting “core socialist values” and “healthy content.” Penalties for AI-related offenses, initially proposed, have been removed, indicating the government’s supportive stance on the technology’s innovative development.
Chinese tech giants like Baidu and Alibaba have yet to publicly launch their ChatGPT-style services, while OpenAI’s ChatGPT and Google’s Bard models remain unavailable in China. The new AI policy mandates that AI service providers should conform to these core socialist values and avoid generating content that could disrupt state power, endanger national security, or incite secession, among other stipulations.
Integrating Self-Censorship in AI Services
Under the new regulations, AI models and chatbots developed in China should integrate functions that ensure generated content is free of harmful information. The law also emphasizes the creation of innovative AI applications that promote positive, healthy content. For instance, 360 Security Technology’s chatbot features a self-censorship function that ends the conversation if a user inputs any “sensitive word.”
The Implications of China’s AI Policy for Developers
Chinese regulators expect AI service providers to design and select training data that avoids any discrimination based on ethnicity, faith, country, or region. Generative AI service providers are also required to report and halt illegal activities if they find users creating illegal content via AI on their platforms. The law further mandates restrictions to prevent minors from becoming addicted to AI services.
While this policy primarily seeks to regulate public-facing AI services, the Shanghai-based lawyer, Xia Heilong, points out that the regulations will also apply to AI services intended for industrial or internal corporate use.
AI Apps Removal from Apple’s China App Store
Following China’s announcement of the new AI policy, multiple generative AI apps were removed from Apple’s China App Store. The action was a response to the new regulation’s requirement that AI apps operating in China secure an administrative license. The removal notice served to OpenCat, a native ChatGPT client, cited “content that is illegal in China” as the reason for the app’s removal.
Conclusion
China’s new AI policy represents a significant move in the regulation of generative AI services, reflecting the government’s attempt to balance innovation with potential risks. While it encourages the innovative use of AI technology, it also raises new compliance challenges for developers and service providers, particularly with the requirement of an administrative license. As the world watches the unfolding of these regulations, the impact of China’s AI policy will undoubtedly shape the global discourse around AI regulation and its future.
Source : Techcrunch
