China’s internet regulator, the Cyberspace Administration (CAC), has unveiled draft measures to manage productive artificial intelligence services and says it asks firms to submit security assessments to authorities before offering their services to the public.“It will be compatible with socialist values”
With the rules prepared within the scope of the draft law called “Administrative Measures for Productive Artificial Intelligence Services”, a standard is tried to be brought in artificial intelligence products. The CAC says China supports AI innovation and applications, but requires the use of secure software, tools and data sources. The agency also underlines that the content produced by productive AI must be compatible with the country’s core socialist values.
The prepared guidelines also state that firms will be responsible for the legitimacy of data used to train productive AI products, and must take measures to prevent discrimination when designing algorithms and training data. However, it is also said that providing real credentials will be mandatory in order to use productive AI tools.
It will come into effect this year
The new rules also require productive AIs not to create content that will embarrass the government, not engage in terrorist or extremist propaganda, engage in racism, or contain other content that could disrupt the economic and social order.It is also stated that if all these rules are not followed, service providers will be fined, their services will be suspended and even face criminal investigations. CAC also says that if productive AI produces inappropriate content, the developer company must update the technology within three months to ensure it doesn’t happen again. For now, these rules are in the form of a draft and are expected to take shape by 10 May. It is planned that the rules will come into effect within the year. Meanwhile, Chinese tech giants such as Baidu, SenseTime and Alibaba are working on artificial intelligence tools like ChatGPT.
42268
