China adopts interim AI rule requiring technology to back the communist system

China on Tuesday implemented a new interim regulation requiring providers of artificial intelligence services to support the communist system and prevent the technology from subverting state power, according to the text of the administrative measure.

The restrictions highlight a renewed emphasis on security under Chinese President Xi Jinping that regulates all technologies and information in the country from perceived threats to Communist Party’s rule.

China currently lists over 100 artificial intelligence companies capable of producing services similar to the American ChatGPT and OpenAI chatbot services and has vowed to become a global leader in the technology in the coming decades.

Artificial intelligence is a new technology that utilizes advanced software and high-powered computers to create text, images, audio and video that can mimic some human brain functions.

The new Chinese interim regulation was published July 13 and went into effect Tuesday. It was approved by China’s Cyberspace Administration and endorsed by seven other agencies, including the Ministry of Public Security, the main police and security ministry.

A central section of the regulation states that those providing AI services must “adhere to the socialist core values, and must not generate content that incites subversion of state power and overthrowing the socialist system.”

Socialist core values are those outlined in the Chinese constitution that describes the country’s brand of communist ideology as Marxism-Leninism with Chinese characteristics.

Chinese AI providers that “possess public opinion attributes or social mobilization capability” are required to undergo security screening, the regulation states.

Other restrictions include prohibitions vaguely described as any AI that “endangers national security,” damages the national image, incites succession, or undermines national unity and social stability. AI also cannot be used to promote “false and harmful information” banned by the Party.

Matt Sheehan, a China expert at the Carnegie Endowment for International Peace, stated in a report China is producing some of the world’s most detailed regulations on AI, including control over algorithms that are the most visible form of AI deployed on the internet.

China in 2021 restricted algorithms for AI requiring registering the software with authorities, and in 2022 added rules for synthetically generated content, before issuing the draft regulation on July 13.

“Information control is a central goal of all three measures, but they also contain many other notable provisions,” Mr. Sheehan said.

The rules seek to bar excessive price discrimination, protect workers’ rights and restrict the use of personal data or using AI for “illegal” purposes.

Senate Majority Leader Chuck Schumer earlier this year announced plans for U.S. regulation of AI and said China could not be allowed to “write the rules of the road” for the cutting-edge technology.

AI has numerous military and intelligence applications that are being developed by both the Pentagon and the Chinese military. Advanced artificial intelligence can be used for intelligence gathering, cyber spying and operating autonomous weapons systems.

The new Chinese regulation was signed by the leaders of major Chinese information and technology ministries, including Minister of Public Security Wang Xiaohong, a close political ally of Mr. Xi.

Mr. Sheehan said the interim law is part of a process of producing a more comprehensive set of regulations to be put in place later this year or in 2024. The main goal of the new regulations is to shape technology “so that it serves the [Chinese Communist Party’s] agenda, particularly for information control and, flowing from this, political and social stability,” Mr. Sheehan stated.

Mr. Sheehan said Mr. Xi does not appear to be as directly involved in regulating AI rules as he has been in other areas. The Chinese president stated in a 2019 speech that China was committed to making AI safe, secure, reliable and controllable.