Understanding China’s New Regulations on Generative AI

Posted by Written by Yi Wu Reading Time: 7 minutes

China is seeking to regulate generative AI alongside policy efforts to foster an enabling environment for AI-led innovation, science and technology, and R&D applications across all industries. The goal behind the recently released draft Measures for Generative Artificial Intelligence Services is to align industry outcomes with China’s core socialist values, social morality, and public order. In other words, balance AI sector innovation with ethical considerations.

The Cyberspace Administration of China released the draft “Measures for Generative Artificial Intelligence Services” (《生成式人工智能服务管理办法(征求意见稿)》, hereinafter, “draft Measures”) in April. Designed to manage generative artificial intelligence (AI) products like ChatGPT, the measures will lay out the ground rules that generative AI services must follow, including the type of content these products are allowed to generate. The draft Measures also highlight issues that are of particular concern to the Chinese government regarding the use of generative AI, such as content moderation, information distortion, and abuse, algorithmic bias and prejudice, and transparency.

What is generative AI?

Generative AI uses deep learning techniques and large datasets to create similar content like images, text, and music. OpenAI’s ChatGPT, a language model capable of generating responses, has gained immense popularity with around 100 million users in just two months.

This success has motivated major internet companies to expedite their own generative AI service platforms. Google launched the “Bard” chatbot, while Microsoft introduced Azure for conversational AI. Chinese tech companies are also heavily investing in developing their generative AI models. Baidu launched “ERNIE Bot,” Alibaba released “Tongyi Qianwen,” and Tencent, 360, Huawei, and other giants have announced their entry into generative AI services. The rapid development has prompted China’s policymakers to consider regulations for this sector.

Understanding China’s generative AI regulations

In recent years, China has emerged as a frontrunner in AI regulations, implementing various initiatives to foster and supervise the growth of the AI sector. Examples include Made in China 2025, the Action Outline for Promoting the Development of Big Data (2015), and the Next Generation Artificial Intelligence Development Plan (2017).

China has also been proactive in enacting laws that govern the ethics of AI businesses and algorithms. As part of their broader efforts to regulate the technology industry, it is conceivable that the Chinese government may impose regulations on AI-based language models, such as ChatGPT.

Getting back to the draft Measures, it’s possible that the regulation can be passed and come into effect later this year and will work in tandem with China’s existing regulatory frameworks. These include laws and regulations like the Cybersecurity Law, Data Security Law, Personal Information Protection Law (PIPL), Deep Synthesis Provisions, and Administrative Provisions on Recommendation Algorithms in Internet-based Information Services. Together, they provide developmental guidance for the AI industry.

In general, the draft Measures seek to reduce the risks and misuse of AI technologies while encouraging AI innovation that is in line with China’s core socialist values, social morality, and public order. By integrating these regulations, China aims to strike a balance between fostering AI advancements and ensuring responsible and ethical AI practices within the country.

Further, and notwithstanding the requirements introduced by the draft Measures, the text states that the government encourages the indigenous development of and international cooperation in relation to generative AI technology and encourages companies to adopt “secure and trustworthy software, tools, computing, and data resources” to that end.

Scope of application

The draft Measures define generative AI as technologies that generate text, pictures, sounds, videos, codes and other content based on algorithms, models, and rules. They specifically regulate the broader category of generative AI technologies, as opposed to merely “deep synthesis” technologies. If enacted, the draft Measures will regulate the activities of service providers offering their products to users in mainland China, regardless of whether the service is based locally or abroad.

Governance of providers

The draft Measures define a “provider of generative AI” as organizations and individuals who utilize generative AI products to offer services, such as chat and text, image, and sound generation. This includes those who facilitate content generation through interfaces and platforms. The draft Measures do not differentiate between back-end technology providers and application-level service providers. Both are held accountable for the content produced by generative AI products and are required to adhere to the PIPL of China to safeguard personal information.

The draft Measures intend to regulate generative AI technologies by imposing direct obligations on the “Providers.” Providers are responsible for ensuring the legitimacy of the data used to train generative AI products and the generated content. They are required to submit security assessment reports to the relevant authority and follow procedures for algorithm filing, modification, and cancellation before offering services to the public using generative AI products. The draft Measures establish clear requirements for service providers, emphasizing their responsibility for the generated content to ensure its truthfulness and accuracy.

Basic requirements for generative AI in China

The draft Measures for Generative Artificial Intelligence Services reinforce existing legal requirements and impose additional obligations on providers of generative AI products and services. These requirements include:

  • Implementing measures to prevent discrimination based on factors, such as race, ethnicity, religion, nationality, geography, sex, age, occupation, etc.
  • Respecting the intellectual property rights of others and refraining from using algorithms, data, and platforms for unfair competition.
  • Taking measures to ensure the authenticity of generated information and avoiding the generation of disinformation.
  • Safeguarding input information and user history, prohibiting the illegal retention of user identity-revealing input information, creating user portraits based on such information, and sharing user input information with third parties.
  • Clearly stating the scope of users, occasions, and usage of their services, and implementing measures to prevent addiction.
  • Providing necessary information to the Cybersecurity Administration of China (CAC) upon request, including dataset sources, size, nature, quality, labeling rules, labeled dataset details, basic algorithms, and technical systems.
  • Suspending or ceasing services if they are used in ways that violate commercial or social ethics, such as engaging in online hype, malicious posting, and commenting, spam creation, writing malicious software, or engaging in improper commercial marketing.
  • Requiring users to provide their real identities and related information, while refraining from generating false information.

Rights of users

When personal information is involved in the services provided, generative AI providers are legally obligated to fulfill the responsibilities of a personal information processor and ensure the protection of personal information. They must establish reporting systems for users to handle individual requests related to modifying, deleting, and blocking personal information. If the generated content infringes upon portrait rights, reputation rights, personal privacy, or business secrets of others, the providers must cease generating such information.

Legal liability

Violations of the Measures will be subject to penalties in accordance with relevant laws. If there is no specific provision for the penalty, the CAC and other applicable regulators may issue warnings, require corrections within a specified timeframe, order the suspension or termination of services, and impose fines ranging from RMB 10,000 to 100,000 (US$1,470 to US$14,700). In cases where inappropriate content is generated, providers must update their technology within three months to prevent its recurrence. These measures aim to ensure compliance, protect personal information, and address content-related issues promptly.

China’s general take on AI technologies

China has been actively promoting the development of AI through various policy initiatives. In 2017, the State Council issued the “Next Generation Artificial Intelligence Development Plan (2017),” outlining strategic objectives for AI development.

In 2023, the Ministry of Science and Technology launched a special deployment of “AI for Science” to accelerate innovation and promote the high-level application of AI in key industries. AI was identified as a key industry in China’s controversial Made in China 2025 industrial plan in which it was stated that China’s goal was to become a global leader in the field by 2030.

Furthermore, the government encourages the development of generative AI technology, both domestically and through international cooperation. Companies are encouraged to adopt secure and trustworthy software, tools, computing resources, and data resources to support their efforts in this field. These measures highlight the government’s commitment to fostering a secure and reliable AI ecosystem while promoting innovation and collaboration in generative AI technology.

Governments at local levels also dedicate efforts to boost AI development. Major cities, such as Shenzhen and Shanghai, have promulgated their own regulations. One highlight of the Shanghai AI Regulation is that it stipulates a certain degree of tolerance for minor infractions to encourage the exploration of scientific frontiers and inspire innovation.

China has made significant strides and gained a competitive edge in various aspects of the AI industry. The country has shown strength in terms of academic papers, patents, cross-border investments, and global AI funding.

According to forecasts, China’s AI investment is projected to reach a substantial amount of US$26.7 billion by 2026. This investment is estimated to account for approximately 8.9 percent of the global AI investment, positioning China as the second-largest destination for AI investment worldwide. These figures indicate China’s strong commitment and growing influence in the field of AI, both domestically and on a global scale.

Comparison with US and EU

China is not the only country concerned with the development of generative AI; others are also expediting their own policymaking processes. The National Telecommunications and Information Administration (NTIA) of the US Department of Commerce has initiated a formal public request for input on policies that should shape an AI accountability ecosystem. The request, open until June 10, 2023, seeks feedback on various aspects, including data access, measuring accountability, and sector-specific approaches to AI in industries, such as employment and healthcare.

In Europe, the UK government recently announced its decision not to establish a new regulator for AI governance, aiming to avoid overly restrictive regulations that could hinder innovation. Instead, existing regulators are tasked with developing their own approaches to AI governance based on the specific needs of their sectors.

On another note, Italy banned ChatGPT, citing concerns related to privacy issues, in March. These examples demonstrate the diverse approaches in addressing AI governance and the varying concerns and considerations surrounding the technology.

Business prospects in China

The AI industry presents significant investment prospects in China and has become a strategic focus for the government. With rapid growth and advancements, China is positioning itself as a global leader in AI technology. Despite some regulatory restrictions and challenges, the country remains committed to fostering a favorable environment for foreign investment and talent in the AI sector.

China’s determination and supportive policies have marked it as an attractive destination for AI investments. With a focus on self-sufficiency in AI technology, China’s commitment to fostering innovation and its favorable regulatory environment create a conducive business environment. The country’s large market size, growing consumer base, and advanced infrastructure also offer ample opportunities for companies to develop and scale their AI products and services.

To fully capitalize on the prospects in China’s AI industry, companies should leverage China’s supportive policies to optimize their business strategies.

By aligning with the government’s objectives and taking advantage of available incentives, investors can unlock various business opportunities. At the same time, they need to stay agile to adapt to the changing business and political climate to effectively navigate the intricacies of the AI market.

About Us

China Briefing is written and produced by Dezan Shira & Associates. The practice assists foreign investors into China and has done so since 1992 through offices in Beijing, Tianjin, Dalian, Qingdao, Shanghai, Hangzhou, Ningbo, Suzhou, Guangzhou, Dongguan, Zhongshan, Shenzhen, and Hong Kong. Please contact the firm for assistance in China at china@dezshira.com.

Dezan Shira & Associates has offices in Vietnam, Indonesia, Singapore, United States, Germany, Italy, India, and Russia, in addition to our trade research facilities along the Belt & Road Initiative. We also have partner firms assisting foreign investors in The Philippines, Malaysia, Thailand, Bangladesh.