1. China has released draft measures for the management of generative artificial intelligence (AI) services, which apply to the research, development, and use of products with generative AI functions, and to the provision of services to the public within mainland China.
2. The measures require that content generated through the use of generative AI reflects Socialist Core Values and does not contain subversion of state power or other harmful content. Providers are also required to respect intellectual property rights and commercial ethics.
3. Providers using generative AI to provide services such as chat, text, image, or audio generation bear responsibility as the producer of the content generated by the product and are required to submit a security assessment before providing services to the public.
The article presents the Measures for the Management of Generative Artificial Intelligence Services, which aim to promote the healthy development and standardized application of generative AI in China. The translation is provided by a group of individuals and edited by Graham Webster. However, an alternative translation from China Law Translate was consulted during editing.
The Measures apply to the research, development, and use of products with generative AI functions and services provided within mainland China. Generative AI refers to technologies generating text, image, audio, video, code or other content based on algorithms, models or rules.
The article highlights that the State supports indigenous innovation in foundational technologies such as AI algorithms and frameworks while encouraging the prioritized use of secure and reliable software, tools, computing and data resources. Providers must abide by laws and regulations while respecting social virtue and good public customs.
Content generated through generative AI must reflect Socialist Core Values while avoiding subversion of state power; overturning of the socialist system; incitement of separatism; harm to national unity; propagation of terrorism or extremism; propagation of ethnic hatred or ethnic discrimination; violent, obscene or sexual information; false information; as well as content that may upset economic order or social order.
Providers must adopt measures to prevent discrimination based on race, ethnicity, religious belief, nationality, region, sex age or profession during algorithm design processes such as selecting training data model generation and optimization service provision. They must also respect intellectual property rights and commercial ethics while ensuring that content generated through generative AI is true and accurate.
Providers are responsible for the legality of sources used for generative AI product pre-training data and optimization training. Data used for these purposes must conform to cybersecurity laws while not infringing intellectual property rights. If personal information is involved in data usage for generative AI products pre-training or optimization training purposes consent from personal information subjects shall be obtained.
When human annotation is used in developing generative AI products providers must formulate clear, specific and practicable annotation rules conforming to the requirements of these Measures. Necessary training of annotation personnel shall be conducted while the validity of annotation content shall be spot checked.
Providers must explicitly disclose user groups, occasions and uses for their services while adopting appropriate measures to prevent users from excessive reliance on or addiction to generated content. Providers have a duty to protect information input by users and usage records while not illegally preserving input information from which it is possible to deduce the identity of users.
The article does not present any unsupported claims or missing points of consideration. However, it is important to note that the Measures may have potential biases towards promoting Socialist Core Values and protecting national unity. The Measures may also limit freedom of expression by prohibiting certain types of content generated through generative AI.
Overall, the article presents a comprehensive overview of the Measures for the Management of Generative Artificial Intelligence Services in China. It highlights the importance of abiding by laws and regulations while respecting social virtue and good public customs. However, it is important to consider potential biases towards promoting Socialist Core Values and limiting freedom of expression when analyzing these Measures.