Beijing pushes for AI regulation

A campaign to control generative AI raises questions about the future of the industry in China

Axel Rangel Garcia

Beijing pushes for AI regulation

last update on 07 May 2024

China’s internet regulator has announced a campaign to monitor and control generative artificial intelligence. The move comes amid a bout of online spring cleaning targeting content that the government dislikes, as well as Beijing forums with foreign experts on AI regulation. Chinese Premier Li Qiang has also carried out official inspection tours of AI firms and other technology businesses while promising a looser regulatory regime that seems unlikely.

AI has been a focus of the Chinese state since 2017 when the State Council laid out a plan to become a world leader in the field by 2030. But the rush of global interest in the technology in the last year, driven in large part by the publicity of generative models such as ChatGPT, has spurred worries among officials that China is falling behind US competitors and that AI-generated content could overrun the country’s controlled internet environment.

One of the concerns is that generative AI could produce opinions that are unacceptable to the Chinese Communist Party (CCP), such as the Chinese chatbot that was pulled offline after it expressed its opposition to Russia’s war in Ukraine.

However, Chinese internet regulation goes beyond the straightforwardly political. There are fears about scams and crime. There is also paternalistic control tied up in the CCP’s vision of society that doesn’t directly target political dissidence—for example, crackdowns on displaying so-called vulgar wealth. Chinese censors are always fighting to de-sexualize streaming content and launching campaigns against overenthusiastic sports fans or celebrity gossip.

AI regulation has become another vehicle for the concerns of an ageing political leadership that doesn’t trust the young people who generate most online content. Top leaders taking a direct interest in an industry like AI is a double-edged sword for Chinese businesses. On the one hand, it grants a publicity and funding boost. On the other, it means inspection tours by officials who likely don’t know much about the field but are keen to be seen to be doing something.

AI's chaotic sprawl sits uncomfortably with the CCP's view of the internet as a walled, well-managed garden.

China's tech successes in the 2000s and early 2010s were helped by authorities who tended to let the market determine winners, then to back the victors. Under Chinese President Xi Jinping, though, political backing has become vital—especially since the crackdowns on the tech industry that began in 2020 and wiped more than $1.1tn in value off the market.

The new regulations are particularly concerned about scamming, a problem that has attracted much attention in China in the last two years thanks to a rash of deepfake cases within China and the kidnapping of Chinese citizens to work in online scam centres in Southeast Asia. Like other buzzwordy tech trends, AI is full of grifting and spam, but scammers and fakes are already part of business in China.

The political leadership's interest in AI offers some potential to rip the government off, especially given that many officials barely understand the field. Take the AI textbooks for preschoolers or simply attach "AI" to any existing business in the hope of getting a grant. In China, businesses often try to jump on the bandwagon of government-approved buzzwords—and Beijing is very unwilling to admit its own mistakes.

However, the fate of cryptocurrency in China offers a more hopeful tale for regulators. Initially, there was a major boom in the industry, mostly driven by money laundering: Chinese traders could buy cheap electricity to waste on crypto mining in yuan and sell the resulting product in dollars.

The Chinese government toyed with the idea of getting in on digital currency but instead cracked down. As a result, Chinese tech and finance firms generally stayed away from crypto, while Western firms invested in useless products and failed ambitions.

If China can pull off focused AI investment while steering clear of the scams that plague the field, it could hit its 2030 target. It's possible that Beijing's lack of effective intellectual property enforcement could even make it an amenable environment for AI development, since copyright lawsuits will likely impose major costs on the industry in the West. But AI's chaotic sprawl sits uncomfortably with the CCP's view of the internet as a walled, well-managed garden.

+ / -