China has introduced provisional rules for generative AI, including a licensing regime for service providers. These regulations come as the use cases for generative AI continue to rise, and China aims to take a leading role in defining the appropriate use of this rapidly evolving technology.
The rules, unveiled by China’s top cyberspace regulator on Thursday, apply to generative AI services, including API providers, that cater to China-based users. However, there are concerns about whether these strict rules will hinder innovation. The policymakers address these concerns by emphasizing that the regulations aim to balance development and security.
One of the primary requirements outlined in the rules is that generative AI providers must uphold core socialist values, which means prohibiting content related to pornography, terrorism, racism, and anything that threatens China’s national security. Additionally, algorithms that have the potential to influence public opinion must be registered with the relevant authority.
To ensure user protection, the rules state that algorithms must not discriminate based on factors such as ethnicity, gender, age, occupation, or health. They should also not be utilized for anti-competitive behavior. Service providers are encouraged to implement anti-addiction systems for underage users, similar to those found in video gaming.
Service providers bear the responsibility of identifying and halting the generative process in case of illegal content. They must then correct the algorithms and report the incident to the appropriate authority. This means that individuals involved in prompts for an image generator or chatbot could potentially face legal trouble.
Regulators are also given the right to access specific details about generative AI models, including their training data, size, type, tagging rules, and algorithms used.
Furthermore, China’s AI development has been primarily driven from the top-down. As a result, the authorities are calling for the establishment of a public data training platform and the sharing of computing power. In Beijing, concrete rules have already been proposed for a state-backed, centralized platform that allocates public cloud resources based on customer needs.
Similar to other critical industries, China is advocating for self-reliant innovation in AI algorithms, frameworks, chips, software platforms, and infrastructure. However, they also encourage equal and mutually beneficial international cooperation in this field.
China’s introduction of provisional rules for generative AI, including a licensing regime, demonstrates their commitment to shaping the future of this technology within their society. While concerns about stifling innovation persist, the regulations aim to strike a balance between development and security. By prioritizing user protection and public interest, China hopes to establish a responsible and controlled framework for generative AI services.
