Chinese flag over keyboard

China moves to regulate generative AI

Image credit: Dreamstime

China's cyber-space regulator has unveiled draft measures that would make companies responsible for the data used to train generative artificial intelligence (AI) models, such as Midjourney and ChatGPT.

China's Cyberspace Administration of China (CAC) has published a proposed set of rules to govern the use of generative AI tools, as the country's tech giants begin to develop their own intelligent chatbots. 

The CAC’s draft measures describe the type of content these products can generate and identify those who should be held responsible for their mistakes.

The CAC said that China supports AI innovation and application and encourages the use of safe and reliable software, tools and data resources. However, it specifies that content produced by generative AI has to be in line with the country's core socialist values and should not subvert state power.

Moreover, companies should ensure the data being used to train these AI models will not discriminate against people based on aspects such as ethnicity, race and gender, the CAC said. They should also not generate false information, the regulator added.

In addition, China has requested firms to submit security assessments to authorities before they launch their offerings to the public. It is also said to mandate service providers to require users to submit their real identities and related information to use the service.

Providers will be fined, have their services suspended or even face criminal investigations if they fail to comply with the rules, the CAC has said. If inappropriate content is generated by their platforms, the agency will require companies to update the technology within three months to prevent similar content from being generated again. 

The publication of these draft measures comes as China's own firms are beginning to launch their own generative AI alternatives, while many other governments are also contemplating how to mitigate the dangers of the technology. 

ChatGPT is estimated to have gained 100 million monthly active users in the first two months after its launch, making it the fastest-growing consumer application in history, according to a UBS study published last month. However, OpenAI has been clear that the software remains a work in progress, as some of the chatbot’s responses have proven to be inaccurate.

In recent weeks, Baidu, SenseTime and Alibaba have shown off in recent weeks their new artificial intelligence models, while Microsoft and Google Bard have already incorporated this technology into their search engines.

Earlier this month, notable technology figures including Elon Musk and Steve Wozniak signed an open letter warning that AI labs were locked in an “out-of-control race” and calling for a six-month pause on all large-scale AI experiments. 

The rise in popularity of these technologies has raised privacy concerns. While some countries such as Italy have chosen to temporarily block these tools, others such as the UK have begun designing more ‘light touch’ regulatory frameworks regarding the safe use of AI

The Chinese public can comment on the proposals until May 10 and the measures are expected to come into effect in China sometime later in the year, according to the draft rules. 

Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.

Recent articles