China to regulate generative artificial intelligence services from August

China has approved a provisional regulation to regulate the services of artificial intelligence generative similar to ChatGPTwhich use models and algorithms to create content such as text, images or videos.

The regulations, which will enter into force provisionally on next August 15establishes the principles, obligations and responsibilities of the providers and users of these services.

According to the text published this Thursday jointly by the Ministry of Science and Technology, the Cyberspace Administration and five other organizations, generative artificial intelligence services will be subject to “current regulations on information security, protection of personal data, intellectual property and scientific and technological progress”.

In addition, they must respect “fundamental socialist values” and “social morality and professional ethics”, and they will be prohibited from “generating content that threatens national security, territorial unity, social stability or the legitimate rights and interests of other people.” ”.

The regulations also require the providers and users of these services to adopt “effective measures” to avoid discrimination based on nationality, ethnicity, religion, origin, gender, age or profession in the algorithm designthe selection of data, the model generated and the service offered.

Likewise, these services must guarantee “transparency and reliability”, identifying as such the content generated by said technology, an initiative that some Chinese applications such as Douyin (the Chinese version of the video service tiktok) had already begun to apply in recent months.

The material generated by artificial intelligence must be “true” and “accurate”, stipulates the regulation, which also announces measures to “avoid the creation of false information”.

In recent weeks, various artificial intelligence projects presented by technology companies such as Baidu, Huawei or Alibaba have aroused great interest in the Asian country, to the point that the official press warned of a possible “bubble” in the market due to “excessive enthusiasm” about this technology.

Last April, the Cyberspace Administration already published the draft of a regulation for the artificial intelligence sector, which required that the content created by ‘chatbots’ and other generative models “reflect fundamental socialist values” and do not “undermine national unity”, “subvert state power” or “incite to divide the country”.