A woman walks past a building whose exterior wall was painted to look like a circuit board with an AI logo in Hangzhou, China, Oct. 30. AFP-Yonhap
Beijing is stepping up its defense of classic literature against AI misuse that it says could undermine and endanger China’s leading ideology.
The Chinese government says AI’s “peculiar adaptations” from classic television dramas based on Chinese literature are “highly deceptive.” Like memes, these edited video clips are made largely for laughs and social media traction, and highlight Beijing’s challenges in regulating artificial intelligence.
The National Radio and Television Administration (NRTA), the agency that oversees state broadcasting, last week issued a notice calling on its supervisory bodies to clean up videos that use AI to create jokes derived from Chinese cultural touchstones.
It said there was an increasing trend of altering video clips from television shows based on novels hailed as the greatest works of Chinese literature, including Dream of the Red Chamber and Journey to the West.
The broadcast watchdog said these AI-remastered jokes “seek to gain traffic without boundaries and disrespect classic intellectual properties.”
It added that they “challenge traditional cultural perceptions, go against the core spirit of the original works and may constitute copyright infringement.”
The AI-remastered videos, which are largely characterised by a humorous or sarcastic tone, can still be found on Chinese social media platforms, with some clearly labelled as generated by AI while others are not.
In an example, characters from the Journey to the West, a 16th century novel based on a Buddhist monk’s journey to India in the seventh century, are happily riding motorcycles on their supposedly arduous journey.
The NRTA notice reflects one of the latest challenges in AI governance in China, which seeks to balance the priorities of using the technology to boost the economy while combating deepfakes to eradicate misinformation and handle the potential impact on public perception.
Traditional culture, in particular, is key to the Communist Party’s governance philosophy, which it sees as part of its cultural rejuvenation campaign.
Zhang Zewei, founder of AI firm Super Brain which specializes in digital replicas of the dead, has footage of him taken by a colleague as he demonstrates the creation of a video using artificial intelligence in Jiangyin, China, Nov. 8, 2023. AFP-Yonhap
The Cyberspace Administration of China, the country’s top internet watchdog, has repeatedly pledged to clean up video content that could be harmful to the young.
The call to clean up AI-remastered content aligns with China’s campaign to control online information and clean up information it deems undesirable, such as that deemed contrary to traditional values or related to materialism, extreme individualism and “historical nihilism.”
Historical nihilism is a term coined by the Communist Party for challenges to its official version of history, which it sees as undermining its legitimacy.
Gu Yu, a Beijing-based lawyer, said such AI videos had not caused significant harm to the public but could influence people unfamiliar with the original works.
“While new technologies should be approached with a degree of tolerance to encourage public innovation, it is also crucial to take action against works that distort basic values,” he said.
But China’s AI regulation is still in an “exploratory phase,” with regulation and lawmaking lagging behind the technological advances, he said.
According to Gu, it is difficult to draft comprehensive laws without sufficient legal precedents, so measures are introduced when problems arise.
Beijing has taken many efforts to regulate AI, including its “Administrative Provisions on Deep Synthesis in Internet-based Information Services” implemented in January 2023, which require clear labelling of content that could confuse or mislead the public.
In August last year, China took measures to manage generative AI services and issued guidelines for multimedia content labelling standards.
The NRTA’s request to clean up AI-remastered content was a trending topic in China this week, ranked second among hot topics discussed on the social media platform Weibo microblog. The tag attracted more than 43.4 million views and over 14,000 discussions.
Many have expressed support, saying regulation is necessary and without oversight such content could lead to severe information pollution.
Some parents also expressed concerns that AI-made jokes might lead to a loss of respect for traditional Chinese classics.
One using the name Mr Bald wrote on Monday that if not regulated “junior school students haven’t even read the Four Great Chinese Classical Novels, yet they are watching funny videos of the Monkey King [the protagonist of Journey to the West] fighting with Jia Baoyu [protagonist of Dream of the Red Chamber].”
However, other online commentators said officials were “overseeing too much,” with one saying “work is already exhausting, and now even our freedom to enjoy entertaining videos after work is being taken away. Do they really have to regulate everything?”
Alex Roberts, who specialises in games and interactive entertainment at law firm Linklaters in Shanghai said: “Chinese authorities are acutely aware of the challenge of balancing the promotion of AI development to bolster growth in industry and the wider economy, while protecting the creative interests of traditional content producers and wider public interests and social values.”
He said Chinese authorities were not alone among global regulators in navigating this challenge and many governments are also exploring ways to regulate AI-generated content.
“But the size and importance of the digital ecosystem in China puts added weight on getting the balance right quickly and effectively,” Roberts said.
Read thefull story at SCMP.
