China plans new AI regulations after Alibaba, Baidu, Huawei launch tech

"AI is a challenge for global governance," says a regulations expert.
Baba Tamim
Stock photo: AI regulations.
Stock photo: AI regulations.


The Cyberspace Administration of China (CAC), China's internet regulator, proposed rules to govern artificial intelligence (AI) tools like OpenAI's ChatGPT on Tuesday.

CAS’s move comes right after the country's two largest tech companies, Baidu and Alibaba, debuted their AI bot tech to compete with the U.S. market.

“China supports the independent innovation, popularization and application and international cooperation of basic technologies such as AI algorithms and frameworks,” CAC said in the draft regulation published on its website.

“It also encourages the priority use of safe and reliable software, tools, computing, and data resources.”

The CAC would hold businesses accountable for the material produced by the services and compel them to undergo a government security evaluation before offering AI services.

According to the rules, content produced by these services cannot include any components that could undermine the state's authority, encourage secession, or disturb the social order.

The proposed regulations are more specific than the generic guidelines being debated in other jurisdictions, noted a Wall Street Journal (WSJ) article on Tuesday.

AI is a challenge for global governance

The proposed regulations state that businesses would be in charge of safeguarding user data and that data used by AI product developers to train their systems must adhere to Chinese legal requirements.

However, despite the US restrictions to purchase advanced semiconductors necessary for training AI models and China's strict censorship regulations, Chinese internet giants like Alibaba, SenseTime, Baidu, and Huawei are moving forward with plans to integrate AI into their services.

Alibaba released Tongyi Qianwen, a large language model it plans to integrate into products such as its search engine, voice assistant, entertainment, and e-commerce.

A ChatGPT-like service called SenseChat and a collection of apps built on SenseNova, a sizable AI model system, were both released by SenseTime Group Inc.

Services based on Pangu, a group of sizable AI models that Huawei created in 2019, are available to enterprise clients in various sectors, including finance, pharmaceuticals, and meteorology.

Last month, Baidu, China's Google, announced ERNIE Bot, ChatGPT-like Chatbot, with few complications.

Globally, governments are debating whether and how to regulate the emerging generation of generative AI tools.

China's proposed regulations maybe be more specific, but Italy temporarily outlawed ChatGPT after discovering that the AI chatbot had collected and saved information inadvertently. The Biden administration in the US has started looking at whether restrictions on the tools are necessary.

"AI is a challenge for global governance," You Chuanman from the Chinese University of Hong Kong, a tech regulation specialist, told WSJ. "Governments from different countries should work together to deliver a global standard."

Previously, according to a report by Chinese state-affiliated media, Global Times, certain professionals in the sector have begun using ChatGPT and other AI-generated chatbot tools in their work. However, the Payment & Clearing Association of China on Monday urged practitioners to use these tools "cautiously."

Add Interesting Engineering to your Google News feed.
Add Interesting Engineering to your Google News feed.
message circleSHOW COMMENT (1)chevron
Job Board