China is often seen as an isolated case when it comes to the regulation of AI. The measures taken so far, and presumably those to come, are characterized by the authoritarian nature of the state. The censorship that is present in all laws related to digital services that are accessible to the public is also evident in the regulations that are enacted to govern artificial intelligence. Nonetheless, many of the regulations passed are also valuable for other markets. Let us review the current state of AI regulation in the Asian giant.
A few years ago, from 2021, China was the first country to adopt new legal measures specifically for the field of artificial intelligence. Prior to that, in 2017, the New Generation Artificial Intelligence Development Plan was launched, detailing how the country could become a leader in AI by 2030. Key aspects of this initiative included advancing the technology, nurturing talent and creating a framework for development with ethical safeguards.
To implement this plan, it was necessary to take measures that would provide a framework for the development of the industry. In 2021, a draft set of regulations targeting the algorithms used by recommendation services was announced. At the time, these types of AI tools were among those with the greatest impact on the general public. Developers and operators were required to be transparent, responsible and maintain a sense of fairness in systems that could influence user behavior or public opinion.
These ethical pillars should underpin innovation in AI. The body responsible for overseeing their implementation is the Cyberspace Administration of China (CAC). This authority coordinates the regulation of artificial intelligence and ensures compliance with the law.
The CAC is also tasked with monitoring compliance with measures to prevent the harmful use of synthetic content, particularly images, whether static or in the form of videos. This regulation deals specifically with deepfakes, which must be appropriately labeled as artificial content, and provides for restrictions to prevent harmful use. The CAC is carrying out this work in collaboration with the Ministry of Industry and Information Technology (MIIT).
The legal framework
To understand the role of each institution in the regulation of AI in China, it is important to define the functions of each agency involved. The Ministry of Industry and Information Technology supports the CAC in the application of legislation, especially at the sectoral level. It oversees the industrial applications of AI and controls technological standards.
Another important player in the regulatory field is the Ministry of Science and Technology (MOST), whose direct involvement is crucial for the promotion of innovation in both the public and private sectors. The MOST works with the other two institutions to promote research and development in the field of AI while complying with the regulatory and ethical framework.
These three institutions are involved in the implementation of another important legislative package: internal measures on generative AI. These measures, which came into force in August 2023, aim to control the development and use of generative artificial intelligence services. The basic principles required include the prevention of discrimination, guarantees of transparency, accuracy and reliability, and respect for social morality (as defined by the state).
In practice, these regulations impose several responsibilities on AI providers. They must exercise strict control over AI. This may include responding to illegal content generated by their platforms by suspending the creation and transmission of content, deleting published content and optimizing the models to solve the problem. There are also requirements for the training data for the models: it must come from legitimate sources, respect intellectual property and for personal information there must be appropriate consent. In this context, the measures also stipulate the protection of all user rights with regard to personal data.
An interesting aspect of Chinese legislation that will inevitably impact the future development of this sector is the fact that these rules contained in the Measures on Generative AI only apply to services that are accessible to the general public in China. This means that services developed and used exclusively by companies, public institutions or academic institutions are not covered by these restrictions, nor are applications aimed exclusively at users outside China.
This could have an impact on how Chinese companies develop their AI applications. With other technologies, it is already common for services to be marketed differently at home and abroad. In any case, the regulations in China are expected to have an impact beyond the country's borders, as many leading companies are able to export AI tools to other markets.
