China unveils temporary rules for the use of generative AI

The Cyberspace Administration of China (CAC) has published a set of interim rules to regulate the activities of generative AI, including API providers serving Chinese customers.

Here's What We Know

First of all, the document requires developers to adhere to basic socialist principles. It prohibits the creation and distribution of content that threatens China's national security.

Algorithms that can influence public opinion must be registered with the relevant authority. Providers of generative AI must obtain a licence in accordance with the law, but the document does not specify who is obliged to do so.

AI should avoid discrimination based on ethnicity, gender, age, profession, or health status. Service providers are also advised to create an addiction protection system for underage users.

Developers of generative AI are responsible for detecting and stopping the generation of illegal content. The document obliges them to correct models in a timely manner and report incidents to the relevant authorities. This means that individuals can be held liable for attempting to create illegal content.

In addition, the regulator has the right to know the specifics of the AI model. The CAC can demand disclosure of information about its training data, size, type, labelling rules and algorithms.

The new rules will come into force on 15 August 2023.

Source: TechCrunch.