The pace of legislation for emerging industries is accelerating. As many governments have noticed the risks of ChatGPT, the generative artificial intelligence (AIGC) industry has also ushered in its first regulatory document in China.

On April 11, the Cyberspace Administration of China publicly solicited opinions on the "Draft Measures for the Administration of Generative Artificial Intelligence Services (Draft for Comments)" (hereinafter referred to as the "Draft for Comments"). The "Draft for Comments" has a total of 21 articles, which provide relatively detailed provisions on the generative artificial intelligence industry, including definitions, access qualifications, responsibilities and obligations, and penalties.

"The country supports independent innovation, promotion and application, and international cooperation in basic technologies such as artificial intelligence algorithms and frameworks, and encourages the priority use of safe and reliable software, tools, computing and data resources." In the "Draft for Comments", the support and encouragement attitude towards the generative artificial intelligence industry is clearly stated.

At present, many domestic technology giants including Baidu, Tencent, Alibaba, Huawei, JD.com, 360 and others have announced corresponding layouts in the generative AI industry. At the same time, many countries including the United States, Italy and Germany have noticed the risks brought by ChatGPT and are considering introducing corresponding regulatory measures.

Several scholars interviewed by The Paper said that the draft for comments puts forward higher requirements for the data governance level and data compliance system construction of enterprises. Reports from several securities firms show that the draft for comments is conducive to the healthy development and standardized application of the industry.

Generative AI is systematically regulated for the first time

"This is the first time that the country has carried out systematic supervision specifically on generative artificial intelligence." Zheng Zhifeng, associate professor at the Institute of Digital Rule of Law and Government of Southwest University of Political Science and Law, told reporters regarding the release of the "Draft for Comments."

In his opinion, there are two points in the "Draft for Comments" that are particularly worthy of attention: First, it continues the concept of full-process and full-factor supervision, focusing on both the research and development of generative artificial intelligence and the use of generative artificial intelligence, and establishing the regulatory responsibilities for the entire process, including security assessment, algorithm filing, pre-trained data compliance, manual labeling compliance, and usage supervision, covering various basic elements such as algorithms, personal information protection, data security, and computing resources.

The second is to strengthen the responsibility of providers. Starting from the perspective of risk control, providers of generative artificial intelligence products or services are required to shoulder a series of responsibilities, abide by laws and regulations, respect social morality and public order and good customs, and bear corresponding responsibilities for algorithms, models, personal information, data, content, and users to ensure that the technology is controllable and used for good.

The reporter observed that the draft for comments particularly emphasizes the protection of personal data and information, and mentions the importance of information security many times. Among them, it requires that the provision of generative artificial intelligence must respect the legitimate interests of others, prevent harm to the physical and mental health of others, damage to the right of portrait, reputation and personal privacy, and infringe intellectual property rights. It is prohibited to illegally obtain, disclose, and use personal information, privacy, and trade secrets. "The draft opinion represents the regulatory direction of the regulatory authorities. It basically continues the regulatory thinking of giving equal importance to safety and development, points out the direction for the development of my country's generative artificial intelligence industry, draws a safety line, and allows companies to comply with laws and regulations when developing and providing generative artificial intelligence products or services." Zheng Zhifeng believes.

In this regard, Yang Yanchao, a researcher at the Institute of Law of the Chinese Academy of Social Sciences and director of the Center for Science, Technology and Law, said, "The Draft for Comments is a positive signal for the development of the generative artificial intelligence industry. What everyone is worried about now is not that ChatGPT is not strong enough, but the security risks that come with it. The Draft for Comments uses relatively concise clauses to determine the basic direction of industrial development, ensure information security, and protect the Chinese economy."

"The draft for comments emphasizes achieving regulatory goals from the perspective of data compliance and data governance, and puts forward higher requirements for each company's data governance level and the construction of a data compliance system." Wu Shenkuo, doctoral supervisor at the School of Law of Beijing Normal University and deputy director of the Research Center of the Internet Society of China, told reporters that this also means that improving the data application capabilities of the access entities will form a certain data-level access threshold.

Qi'anxin's security experts told The Paper that the release of the Draft for Comments and the strengthening of global supervision of AIGC mean that data security and privacy protection have become the premise and key to the development of AIGC. As AIGC technology becomes more mature and popular, companies should pay more attention to data security and privacy protection, and actively seek support from professional network security technology and consulting services when conducting related businesses to ensure the security and reliability of their businesses.

May drive demand for data security industry

In Zheng Zhifeng's view, artificial intelligence technology is crucial to the construction of a digital China, and two points need to be focused on: First, the current development of the domestic generative artificial intelligence industry is still in the catching-up stage, and supervision should give more consideration to the goal of promoting technological development; second, generative artificial intelligence has the characteristics of a general technology, which can be widely empowered in medical care, education, entertainment, government affairs, etc., and supervision will need to be refined from specific scenarios in the future.

According to research reports from multiple securities firms, the "Draft for Comments" will accelerate the development of the generative artificial intelligence industry.

According to a research report by Guolian Securities, the "Draft for Comments" is the first regulatory document in the field of generative artificial intelligence. Its release is in line with expectations. It is the specific implementation of superior laws such as the "Cybersecurity Law" in the field of generative artificial intelligence, which is conducive to the healthy development and standardized application of the industry.

"Network security and data security are the key to ensuring the healthy development and standardized application of the generative artificial intelligence industry. All companies related to the industry chain will benefit from the development of the generative artificial intelligence industry to varying degrees." Guolian Securities said that the "Draft for Comments" mentioned that the provision of generative artificial intelligence products or services should comply with the requirements of laws and regulations, respect social morality, public order and good customs; the pre-training and optimized training data used for generative artificial intelligence products should comply with the requirements of laws and regulations such as the "Network Security Law"; providers shall bear the obligation to protect users' input information and usage records during the provision of services. The development of the generative artificial intelligence industry may drive the demand for network security and data security.

Ping An Securities Research Report analyzed that the "Draft for Comments" has made provisions for the generation of content, subject responsibilities, data sources and data processing. In terms of content, it is required not to endanger the country and society, and not to spread false, rumor-mongering and harmful information; no discrimination should be allowed in the process of algorithm design, training data selection, model generation and optimization, and service provision; respect intellectual property rights and ensure the accuracy of generated content; and no infringement of the legitimate rights and interests of others. Organizations and individuals that use generative artificial intelligence products to provide services shall assume the responsibilities of content producers and fulfill the obligations of information protection.

Ping An Securities said that the development of new technologies and regulatory compliance are mutually reinforcing. In recent years, technologies such as mobile Internet, cloud computing, big data, and blockchain have emerged rapidly, and have achieved orderly and vigorous development. Behind this is the improvement of regulatory institutions in the construction of network security mechanisms and the guidance of industrial policies, which has truly realized "technology for good". At the same time, regulatory policies are also actively adapting to the innovation of information technology. Especially since the launch of AIGC, regulators have gradually become clearer about the data processing subjects, data processing processes, and the accompanying national security and information security.

"We expect that as more AIGC technologies are put into use, some potential risks will be more recognized and exposed, and regulatory measures will be more targeted. The Draft for Comments is only the first step."

Many countries strengthen regulation of artificial intelligence

Earlier, the Wall Street Journal reported that the Biden administration has begun studying whether it is necessary to restrict related tools such as ChatGPT. On April 11, the U.S. Department of Commerce officially solicited public opinions on relevant accountability measures, including whether new AI models with potential risks should undergo an approval and certification process before release. This is seen as the first step towards potential regulation of the field of AI.

"my country is on par with the European Union and the United States in the areas of personal information protection and data legislation, and has become a leader in a certain sense in the areas of algorithm governance and deep synthesis governance." Zhang Linghan, professor at the Institute of Data Rule of Law at China University of Political Science and Law, told The Paper.

"The accelerated pace of legislation in emerging fields has become the norm in countries around the world. Both the European Union and the United States have actually proposed governance plans for generative AI within this week." In her view, mastering the right to speak and formulate rules in AI governance and taking the lead in forming new national competitive advantages have become the goals of different legal jurisdictions in the world.

"The explosion of generative AI such as ChatGPT is a favorable opportunity to promote deep synthesis to become a specialized field of AI governance." Zhang Linghan said that the institutional design of the "Draft for Comments" on the one hand continues the existing institutional arrangements on identification, evaluation, and filing in the "Regulations on the Management of Deep Synthesis of Internet Information Services", and on the other hand, it puts forward new compliance requirements for training data, manual labeling, etc. Compared with the "Regulations on the Management of Deep Synthesis of Internet Information Services", it does reflect the new trend of updating the regulatory logic and iterating the institutional system.

She believes that the healthy development of generative artificial intelligence represented by ChatGPT should be promoted, and a governance system that conforms to China's local economic, social, political, and value systems and can promote the long-term and legal development of the country's artificial intelligence technology industry should be actively built, which will help to form a real institutional competitive advantage.

The reporter of The Paper found that many countries have introduced or are discussing regulatory measures on ChatGPT. In early April, Ulrich Kelber, the German Federal Data Protection Commissioner, said that Germany may follow Italy's footsteps and "ban" ChatGPT due to data security issues.

On March 31st local time, the Italian Personal Data Protection Agency announced that the use of ChatGPT would be banned from now on, restricting ChatGPT's developer OpenAI from processing Italian user information and initiating an investigation.

According to Reuters, Tesla CEO and Twitter "head" Elon Musk, together with a group of artificial intelligence experts and industry executives, called in an open letter to suspend the training of artificial intelligence systems more powerful than OpenAI's newly launched GPT-4 model for at least six months, on the grounds that they pose potential risks to society and humanity.