China Issues Draft Regulations on AI Security; Aims to Lead in AI by 2030
Summary:
China has issued draft regulations regarding the security of generative artificial intelligence (AI) services. The rules call for evaluations on data used in AI training and content promoting harmful behaviors will be blacklisted. These restrictions also apply to data under Chinese internet censorship. The companies need to obtain individual consents for using personal data, including biometric information. The recent regulations come amid China's ambitions to become a global leader in AI by 2030.
Draft regulations regarding security measures for firms offering generative artificial intelligence (AI) services have been issued by China, which include limitations on data sources utilized for AI model education. The tentative rules were announced on Wednesday, Oct. 11, by the National Information Security Standardization Committee, which includes delegates from the Cyberspace Administration of China (CAC), the Ministry of Industry and Information Technology, and law enforcement bodies. Generative AI, seen in the achievements of OpenAI's ChatGPT chatbot, develops the capacity to undertake operations through the study of historical records, generating new content like text and images from this instruction.
According to the committee, security assessments should be carried out on the content used to educate publicly available generative AI models. Content with over "5% in the form of illegal and injurious information" is to be blacklisted. This includes content that promotes terrorism, violence, destruction of the socialist system, defamation of the nation's standing and behavior that undermines national unity and social stability. The suggested regulations also stress that data under Chinese internet censorship should not be used to train these models.
This announcement comes just a little over a month after regulators permitted leading Chinese tech companies including Baidu, the search engine leader, to launch their generative AI-integrated chatbots to the public. The CAC has regularly expressed the need for firms to submit security assessment reports to regulatory entities before taking generative AI-controlled services public since April. In July, the cyberspace regulating body issued a series of regulations overseeing these services, which industry observers pointed out were considerably less rigorous compared to the original April draft rules.
The latest draft security provisions mandate that companies engaged in training these AI models gain clear consent from individuals whose specific data, incorporating biometric information, is used for educational purposes. Plus, the rules contain thorough directives on how to avoid intellectual property violations.
Internationally, countries are facing the challenge of devising regulatory norms for this technology. China regards AI as an area where it seeks to match the United States, and aims to establish itself as a worldwide leader in this sector by 2030.
Published At
10/13/2023 10:12:34 AM
Disclaimer: Algoine does not endorse any content or product on this page. Readers should conduct their own research before taking any actions related to the asset, company, or any information in this article and assume full responsibility for their decisions. This article should not be considered as investment advice. Our news is prepared with AI support.
Do you suspect this content may be misleading, incomplete, or inappropriate in any way, requiring modification or removal?
We appreciate your report.