It has been revealed that an official from the Chinese judicial authorities attempted to use OpenAI's ChatGPT to smear Japanese Prime Minister Sanae Takaichi.
On February 26 (local time), OpenAI stated in its report on 'Preventing Malicious Use of AI' that an official from the Chinese judicial authorities planned and documented a coordinated smear campaign against Prime Minister Takaichi using ChatGPT in mid-October last year.
This official, through an operation called 'cyber special operation,' sought to create negative public opinion about Prime Minister Takaichi by posting negative comments, sending emails critical of her foreigner policy to politicians while impersonating foreigners, framing her as an extreme right-wing figure, and inciting anger over U.S. tariffs on Japan. According to the report, he asked ChatGPT to help formulate such a plan after Prime Minister Takaichi publicly criticized the human rights situation in China's Inner Mongolia Autonomous Region in October last year.
OpenAI said that ChatGPT refused the request to provide advice on such operations. However, at the end of October, this official again asked ChatGPT to edit a report containing a similar operation. The company added that it appears the official executed the operation without ChatGPT's assistance.
In fact, on social networking services (SNS) such as X (formerly Twitter), numerous posts and YouTube videos criticizing Prime Minister Takaichi with the hashtag 'right-wing symbiote,' mentioned in the document, were uploaded. However, it was found that these operations had little real impact, as most of the YouTube videos and SNS posts failed to gain traction.
OpenAI also revealed that, in addition, the Chinese judicial authorities have conducted more than 100 tactics, such as creating and distributing fake obituaries and tombstone photos to suppress dissidents, and targeting human rights organizations. In this process, hundreds of thousands of posts appeared on Chinese social media, and tens of thousands appeared on overseas platforms. The company added that Chinese authorities appear to have used not only ChatGPT, but also domestic AI models such as DeepSeek.
Meanwhile, at a regular press briefing, Chinese Foreign Ministry spokesperson Mao Ning stated, "I am not well aware of the content mentioned, nor do I see any evidence to support such claims," according to Bloomberg News.
© The Asia Business Daily(www.asiae.co.kr). All rights reserved.
