China this week announced a list of administrative measures intended to regulate ChatGPT-like AIs, which should be more in line with “ core values of socialism “.
Beijing does not want to curb the development of generative artificial intelligence technologies. On the other hand, it wants to supervise them without delay and make them comply with the law on network security, in force in the Middle Kingdom. To do this, it has just submitted a series of measures for public comment, which Clubic details for you.
China wants AI to reflect ‘values of socialism’
China wishes that all the measures which will be definitively adopted at the end of the consultation period apply as well ” research, development and use of generative artificial intelligence products which will aim to provide services to the population. It is therefore the entire AI chain that is concerned here, whether it generates text, images, sound, video or other.
The fourth proposal made by Beijing becomes more interesting, more political. Generative AI products or services must comply with the ” requirements of laws and regulations, respect social morality, public order and morality “.
Thus, the regime translates its desire for a generative AI that reflects ” the fundamental values of socialism and which does not challenge the socialist system. Incitement to divide the country, to promote terrorism, extremism, ethnic hatred and discrimination, to disseminate information that is obscene, pornographic, false or likely to disrupt the economic and social order, will be prohibited.
The government wants to be able to trace the users of generative AIs
Beijing thus intends to intervene in the design of algorithms and language models, by ensuring that they banish all forms of discrimination (age, sex, creed, profession, ethnicity, etc.), and by enforcing intellectual property rights. , without forgetting business ethics, so that AI does not contribute to offering advantages that would help to implement unfair competition in this or that field.
Point 4 of Article 4 submitted by the Chinese authorities is also interesting. ” Content generated by generative artificial intelligence should be true and accurate, and measures should be taken to prevent the generation of false information “. This is precisely one of the reasons that are pushing the CNIL today to investigate ChatGPT, after the filing of several complaints.
China should also require that any generative AI service undergo a security assessment, to always remain in compliance with the country’s various regulations. And each user of a chatbot will have to provide usable information to identify them, on the grounds of ensuring the nation’s security.
Source : Cyberspace Administration of China
15