Hong Kong CNN
In what the state-run media called China's first criminal case involving the AI chatbot, police in China detained a Chinese man who they claim used ChatGPT for creating fake news and spreading it online.
A police statement in the province of Gansu said that the suspect used ChatGPT for a fake report about a crash and then posted it online to make money. In a statement issued on Sunday, police stated that the article had received around 15,000 views.
ChatGPT developed by Microsoft's OpenAI (MSFT), which is blocked in China. However, users of virtual private networks can access it.
Since 2011, China has been grappling with the issue of train crashes. The authorities were under pressure to explain to the public why they had not provided timely updates about a bullet-train collision that occurred in Wenzhou, resulting in 40 fatalities.
Authorities in Gansu said that the suspect was identified as Hong and was arrested in Dongguan, a city in the southern province of Guangdong, on May 5.
In a statement, the Gansu Police said that Hong had used modern technology to spread false information on the Internet, which was widely distributed.
They added that Hong's actions amounted to provoking quarrels.
The police said that this arrest is the first one in Gansu after China's Cyberspace Administration passed new regulations to curb the use of fake news in January. State broadcaster CGTN reports that this is the first arrest in China of a person who was accused of using ChatGPT for fabricating and spreading fake news.
Deep fake, also known as deep synthesis or deep synthesis is a highly realistic textual/visual content created by artificial intelligence.
Users are prohibited from creating deep fake content in China's heavily-censored Internet. Topics already banned by law. The document also details the procedures to be followed for removing content deemed false or harmful.
The arrest was also part of a 100-day crackdown on internet rumors launched in March by the Internet branch of the Ministry of Public Security.
Baidu (BIDU), Alibaba (BABA), and other Chinese internet giants have been trying to catch up to OpenAI by launching their versions of ChatGPT.
Baidu launched 'Wenxin Yiyan,' or ERNIE Bot, in March. Alibaba released 'Tongyi Qianwen' two months later. This roughly translates to seeking truth by asking 1,000 questions.
China's cyberspace regulator released draft guidelines last month for public comment. It said that generative AI services will be required to undergo a security review before they can operate.
The service providers will be required to provide details on the type and scale of data that they use, basic algorithms, and other technical information.