The incident has users worried about ChatGPT’s capabilities. This tool can be used to spread fake and fake news in a sophisticated way.
On a February day in Hangzhou, Zhejiang Province, China, an advertisement posted on the Internet caused an uproar by announcing that it would end the regulation of road space allocation from March 1st. This policy prohibits people from driving on certain stretches of the road, especially at peak times, using signs.
This notice mentioned three reasons for lifting the ban and emphasized that: “Although the traffic congestion reduction policy has been in effect since 2006, it still causes a lot of inconvenience to the community.” Previously, the Hangzhou city government postponed the road space allocation regulation from December 2022, but said it would apply again in February 2023.
So this government announcement is of interest to many people. However, when local newspapers got involved, they discovered that it was just a fake message, recently created by ChatGPT, the popular chatbot.
On February 17, Hangzhou People’s Radio said that the address of the person who made the confusing announcement was located in a residential area. This person asked ChatGPT to write an article about deregulating road traffic distribution in China. He recorded and archived the entire process and shared a screenshot of ChatGPT’s response on WeChat.
“At this time, the city government has not announced any policy changes,” City Radio said. At the same time, the person who spread the fake news apologized to the entire city and the people in the WeChat group with whom he shared the screen.
According to Sixth Tone, the incident caused an uproar among many users on social networks. They care about the capabilities of smart chatbots like ChatGPT. “The scariest thing is that this ad was written by AI,” reads a widely shared comment on Weibo.
Thanks to its ability to generate realistic responses, ChatGPT has been a big hit in China in recent weeks, although it has yet to be officially launched in the country. Many people have tried to access and copy an “underground” service or even a fake ChatGPT to trick others.
But the potential and sophistication of the technology behind ChatGPT worries many legal experts in China. They warn that this tool can generate new content based on a huge data warehouse, therefore it can become a source of rumors and false information if there are no specific censorship measures in place. Experts expressed concern that the specific legal regulation with many loopholes would make it difficult to manage violations.
“The content written by AI technology is not always correct, and we need expert knowledge to verify that the information provided by ChatGPT is correct,” said Xie Lianjie, internet law attorney at Sixth Tone.
Experts added that smart tools like ChatGPT can teach programmers completely false information to spread fake news and fraudulent websites.
According to Sixth Tone, Chinese lawmakers have taken steps to help people protect themselves against rumors and fake news on the Internet by requiring the source of all information. In 2022, the Cyberspace Administration of China resolved 172 million cases involving the dissemination of unauthorized and inappropriate information, an increase of 3.6% over last year.