On the 13th of July 2023, the Chinese Government published the Provisional Administrative measures on Generative Artificial Intelligence Services.
It is key to note that under Article 1, the following measures have been drafter on the basis of the following legislation; “Cybersecurity Law of the People’s Republic of China”, the “Data Security Law of the People’s Republic of China”, the “Personal Information Protection Law of the People’s Republic of China”, the “Science and Technology Progress Law of the People’s Republic of China.”
The following provisions outlined in the paper shall not apply to the research and development of Generative AI, where they do not provide Generative AI services to the general public.
Article 4 outlines that the use of Generative AI services must comply with the law, administrative regulations and the following provisions:
- Adhere to the Core Socialist Values, and must not produce content prohibited by laws and administrative regulations, such as inciting subversion of state power, overthrowing the socialist system, endangering national security and interests, harming the country’s image, inciting separatism, undermining national unity and social stability, advocating terrorism, extremism, advocating ethnic hatred, ethnic discrimination, violence, obscenity, pornography, and false and harmful information;
- Employ effective measures to prevent discrimination such as on ethnicity, creed, country, region, gender, age, occupation, health, and so forth, in the process of designing algorithms, selecting training data, generating and optimizing models, and providing services;
- Respect intellectual property rights and business ethics, keep commercial secrets, and must not use algorithms, data, platforms, and other advantages to carry out monopolistic and unfair competition conduct;
- Respect the lawful rights and interests of others, must not endanger the physical and psychological health of others, and must not infringe upon others’ rights and interests in portraiture, reputation, honour, privacy, or personal information;
- Based on the characteristics of the type of service, employ effective measures to increase the transparency of generative AI services and increase the accuracy and reliability of generated content
Chapter II- Technological Development
Article 5 and 6 are centred around ensuring the innovation of Generative AI. Article 5 outlines the papers aim to encourage the innovation of Generative AI technology in all fields. This approach is continued in Article 6 in which the intention to encourage independent innovation in basic technologies is stressed.
Article 7 moves on to outline the provisions which Generative AI providers must comply with when conducting data training processing.
(1) Using data and underlying models with lawful sources;
(2) Where intellectual property rights are involved, the intellectual property rights enjoyed by others in accordance with law must not be infringed;
(3) Where personal information is involved, the individual’s consent shall be obtained or other circumstances provided for by laws and administrative regulations shall be met;
(4) Employ effective measures to improve the quality of training data, and enhance the authenticity, accuracy, objectivity, and diversity of training data;
(5) Other relevant provisions of laws and administrative regulations such as the “Cybersecurity Law of the People’s Republic of China”, the “Data Security Law of the People’s Republic of China”, and the “Personal Information Protection Law of the People’s Republic of China”, as well as the relevant regulatory requirements of relevant regulatory departments.
Chapter III- Service specifications-
This chapter concerns the protection of users and their data.
Article 9 states that providers must bear the responsibility of online information content producers. Providers should sign service agreements with any users who register their services.
Article 10 outlines that providers must disclose the uses of their services to ensure users understand and lawfully use generative AI.
Under Article 11, providers must perform their obligation to protect users input information and comply promptly with any users requests for information.
Chapter IV- Supervision and Inspection
Under Article 17, security assessments shall be carried out in accordance with any state provisions where Generative AI services with public opinion attributes are provided.
Where users discover generative AI services do not comply with the following measures they have the right to make complaints to the relevant authority. Under article 19 the relevant regulatory department are then required to carry out inspections of services.
Where providers violate any of the provisions of these measures, the relevant authority is to provide punishments in accordance with the following legislations; ‘Cybersecurity Law of the Peoples Republic of China, Personal Information Protection Law of the Peoples Republic of China, Science and Technology Progress Law of the Peoples Republic of China,’ The sanctions are outlined in Article 21.