The Administration of the Cyberspace of China has published this Tuesday a draft with the rules that must be followed by the developers of tools of artificial intelligence in the country. A strong regulation that, among other aspects, will require the identification of all users of these tools and the content they generate agree with the “socialist values” that make up the propaganda of the Chinese government.
“Content generated by generative artificial intelligence should embody core socialist values and must not contain any content that subverts state power, advocates the overthrow of the socialist systemincite the country division or undermine the National unity”, establishes the Internet regulatory authority in China.
Although he, for its acronym in English, ACC has indicated that it seeks to hear comments on the content of the new regulation, it is unlikely that these imply any change in the norms that the authority of the communist regime has designed and that must enter into force at the end of this year.
The CAC will force developers to pass a security review before authorizing release of their products and must be registered in a database which will be created. Once publicly available, each platform must verify the identity of your users so they can be tracked.
The draft, which is titled “Administrative Measures for Generative Artificial Intelligence Services”, he states that his goal is “the healthy development and standardized application of generative AI technology”. Other measures that he takes are the prohibition of all content with “terrorist or extremist propaganda”, “ethnic hate” or “other content that may disturb the economic and social order”.
The Chinese authority points to objectives that seem unfeasible with the current development of generative artificial intelligences such as the one that do not generate false information. OpenAI has not achieved this with ChatGPT and it is not a prospect that is in the immediate future, while the chatbots announced by some Chinese companies in recent weeks have offered a worse performance than that and a greater tendency to provide wrong content.
This has happened with Erniethe Chinese search engine chatbot Baidu released last month, or Tongyi Qianwenannounced by Alibaba this week. In a demonstration by the company, the AI ”could write poems in Chinese and French and solve basic math problems, but it had problems with simple logic”, according to Financial Times.
daniel zangCEO of Alibaba, declared in the presentation of the Tongyi Qianwen chatbot, before the new CAC regulations were known, that we are “at a watershed technological moment powered by generative AI and cloud computing. In 10 or 20 years, when we look back, we will realize that we were all on the same starting line. Taking advantage of these future opportunities is our common desire and requires a shared vision.” It is probable that the new regulation supposes a Alteration in Alibaba’s plans to bring this AI to all its products, starting with the DingTalk collaborative tool and Tmall Genie smart speakers.
andy chun, a professor at the University of Hong Kong, told the AFP agency that “the CAC draft is one of the strictest measures for generative AI so far.” For this reason, AI companies will need to “take great care to ensure that each data source used for AI learning is within established guidelines, is accurate, unbiased, and does not infringe the intellectual property rights of others. Ensuring accuracy is difficult. No generative AI system to date can do that.”.
Discussion about this post