China eyes at least 50 sets of AI standards by 2026, from chips to applications and safety

The MIIT’s standardisation initiative echoes the China-led resolution in the United Nations, which calls on the international community to create a “free, open, inclusive and non-discriminatory” business environment among wealthy and developing nations for AI development. This resolution on Monday was unanimously adopted in the UN General Assembly.

By establishing standards for the fast-growing artificial-intelligence sector, China is looking to close the gap with the United States in developing the technology and broadly adopting it to raise the efficiency in various industries. Photo: Shutterstock

“AI is the foundational and strategic technology driving the new round of technological revolution and industrial transformation,” the draft policy said. By accelerating AI’s integration within the country’s economy, the MIIT said this would “profoundly change industrial production and economic-development patterns”.

As such, AI would “play an important role” in raising China’s manufacturing capabilities and internet prowess, according to the agency.

The MIIT’s draft policy has taken on a pro-market, soft-law approach – instead of the typical command-and-control regulation – to guide and promote China’s AI industry development, according to You Chuanman, director of the Institute for International Affairs Centre for Regulation and Global Governance at the Shenzhen campus of the Chinese University of Hong Kong.

“It’s an innovation-oriented and market-friendly regulatory approach,” You said. “This is more about enabling and promoting the development of the technology and its ecosystem,” he added, describing it as beneficial to other industries.

More than a quarter of China’s 369 unicorns – start-ups valued at more than US$1 billion – are involved in the nation’s artificial intelligence and semiconductor sectors. Photo: Shutterstock

The MIIT’s draft policy listed a total of 12 areas as critical technologies across the AI supply chain. These include LLMs, natural-language processing, computer vision and machine learning – a subfield of AI that refers to systems used to perform complex tasks similar to how humans solve problems.

China’s AI industry chain, according to the draft policy, consists of four layers: the foundation – including the computing power, algorithms and data required to train LLMs – as well as the framework, model and applications.

In April, Alibaba Group Holding and chairman Joe Tsai said in a podcast interview that China trails the US by two years in the global race to lead AI development, as mainland enterprises struggle with Washington’s tech export restrictions. Alibaba owns the South China Morning Post.

Темы и теги

Китай Наука и Технологии

комплект чип искусственный искусство стандарт Организация Объединенных Наций miit приложение стандартизация безопасность