Chinese internet regulators are stepping up efforts to combat the misuse of artificial intelligence to impersonate celebrities, which is being used to falsely promote products, gain followers or boost online views, with more than 47,000 pieces of harmful content removed and 3,961 accounts penalized for misleading the public.
The Cyberspace Administration of China, the country's top internet regulator, announced the figures on Friday. It called on internet platforms nationwide to take greater responsibility and adopt more effective measures to prevent AI misuse, protect users' legitimate rights and interests, and maintain a cleaner online environment.
On Oct 18, Chinese actress Wu Yue posted a video on Sina Weibo, saying that a netizen had used AI to fake her face and clone her voice to promote a facial serum. Wu condemned the act, questioning how much money had been scammed from consumers and demanding restitution. The offending account was banned the same day.
On Oct 26, Yi Zhongtian, a professor at Xiamen University in Fujian province, used his WeChat official account to refute AI-generated videos falsely claiming that he had Alzheimer's disease.
On Friday, the administration exposed a number of WeChat accounts that had used AI to fake celebrity images, synthesize voices, or misuse images of public figures and their family members to promote products and attract viewers. It also exposed accounts spreading rumors, including false claims about celebrities' health, and said the accounts had been penalized.
Several other accounts were penalized for impersonating public figures, making false statements, fabricating personal experiences, posting emotional content to lure followers, or using AI to create inappropriate images of celebrities to attract followers and profit from marketing, according to the administration.
Over the past year, Chinese internet regulators have prioritized efforts to combat AI abuse that disrupts online order, launching ongoing campaigns targeting the use of AI to impersonate celebrities and post misleading marketing information or infringe on people's lawful rights and interests in cyberspace.
To further regulate online conduct and strengthen cyber governance, the administration has also launched a two-month campaign to clean up health-related rumors, with a particular focus on accounts spreading illegal financial information.
The administration said the campaign will target online accounts that fabricate welfare-related messages to mislead the public, impersonate medical professionals to spread pseudoscientific claims, or use AI to generate health-related rumors.
It said the crackdown will focus in particular on the use of AI to mass-produce homogenized health-science content in the form of text, images, audio and video to manufacture fake trending topics and promote related products. It will also target the use of AI face-swapping and voice-cloning technologies to steal the images and voices of well-known experts for fake science popularization and livestream selling.
Also targeted are attempts to induce AI models to generate and spread rumors, such as fake "mortality rates" and "survival rates" and fabricated scientific data, in order to sell health supplements, as well as the use of AI to create fake official notices, official health alerts and exaggerated efficacy claims in images or videos to mislead netizens.
Additionally, the authority said it would urge internet platforms to strengthen verification of whether users are health professionals or doctors and improve their ability to detect AI-generated content so that such material can be identified and removed promptly.
caoyin@chinadaily.com.cn
