Man Arrested in China for Misusing ChatGPT to Generate Fake News of Train Crash

Chinese police have detained a man for allegedly generating fake news of a train crash and disseminating it online using artificial intelligence technology to “concoct information” and post it on multiple accounts, in what was reported to be China’s first arrest for misuse of ChatGPT.

Police in northwestern Gansu province said in a statement on Sunday that a suspect surnamed Hong had been detained for “using artificial intelligence technology to concoct false and untrue information”.

The case first caught the attention of the cyber division of a county police bureau when they spotted a fake news article that claimed nine people had been killed in a local train accident on April 25, the Hong Kong-based South China Morning Post reported on Monday.

The cybersecurity officers in Kongtong county found the article simultaneously posted by more than 20 accounts on Baijiahao, a blog-style platform run by Chinese search engine giant Baidu.

The stories had received more than 15,000 clicks by the time it came to authorities’ attention, it said.

The Gansu public security department said Hong was suspected of the crime of “picking quarrels and provoking trouble”, a charge that normally carries a maximum sentence of five years. But in cases that are deemed especially severe, offenders can be jailed for 10 years and given additional penalties, the Post reported.

This is the first time the public has been made aware of an arrest by Chinese authorities after Beijing’s first provisions to regulate the use of “deepfake” technology officially took effect in January, it said.

The police said they traced the origins of the article to a company owned by the suspect Hong, which operated personal media platforms registered in Shenzhen in Guangdong province in southern China. Some 10 days later a police team searched Hong’s home and his computer and detained him.

The statement said Hong confessed to bypassing Baijiahao’s duplication check function to publish on multiple accounts he had acquired. He input the elements of trending social stories in China from past years into ChatGPT to quickly produce different versions of the same fake story and uploaded them to his Baijiahao accounts, it said.

While ChatGPT is not directly available to Chinese IP addresses, Chinese users can still access its service if they have a reliable VPN connection.

Chinese IT outlets were experimenting with their versions of the ChatGPT after Microsoft and Google announced their innovations.

China closely monitors its social media through firewalls, especially Sina Weibo, which has over 592 million users to ensure no critical content against the ruling Communist Party of China (CPC).

China’s top internet regulator has long voiced concern that unchecked development and use of deep synthesis technology could lead to its use in criminal activities such as online scams or defamation.

As ChatGPT has gone viral in recent months, China’s law enforcement agencies have repeatedly voiced suspicion, and even warnings, about the technology.

In one of the first comments on the chatbot made by the Chinese security apparatus, police in Beijing specifically warned the public in February to be wary of “rumours” generated by ChatGPT, the Post reported

---