Welcome to the forefront of conversational AI as we explore the fascinating world of AI chatbots in our dedicated blog series. Discover the latest advancements, applications, and strategies that propel the evolution of chatbot technology. From enhancing customer interactions to streamlining business processes, these articles delve into the innovative ways artificial intelligence is shaping the landscape of automated conversational agents. Whether you’re a business owner, developer, or simply intrigued by the future of interactive technology, join us on this journey to unravel the transformative power and endless possibilities of AI chatbots.
World
World
The AI firm says a Chinese official’s activity on the chatbot pointed to wider “cyber special operations” to intimidate both domestic and foreign critics.
The OpenAI logo is displayed on a mobile phone in front of a computer screen with output from ChatGPT. (File photo: AP/Michael Dwyer)
This audio is generated by an AI tool.
SINGAPORE: OpenAI on Wednesday (Feb 25) said it banned the ChatGPT account of an individual linked to Chinese law enforcement, who had tried to use the chatbot to plan an influence operation against Japanese Prime Minister Sanae Takaichi.
In a periodic threat report, the American artificial intelligence firm said the individual’s “activity revealed a well-resourced, meticulously-orchestrated strategy for covert IO against domestic and foreign adversaries”. This involved several earlier large-scale operations that employed hundreds of staff, thousands of fake social media accounts and local Chinese AI models.
OpenAI has previously identified Chinese influence operations in earlier threat reports, including the documented “Spamouflage” network allegedly pushing pro-China messages while attacking critics. One website from the network was spotted as part of the Chinese law enforcement individual’s ChatGPT activity, said OpenAI in its latest findings.
This user’s attempt to target Takaichi was detected in mid-October, after she criticised human rights in the Chinese autonomous region of Inner Mongolia. The user asked ChatGPT to help with generating negative comments about the Japanese premier; fake complaints about her position on immigration and the cost of living; accusations of Takaichi having far-right leanings; content to stir up anger against the United States over tariffs; and positive sentiments about conditions in Inner Mongolia.
OpenAI said its generative AI chatbot refused to assist, only for the user to return at the end of the month to ask ChatGPT to edit a status report on an anti-Takaichi operation – suggesting the user had gone ahead without using OpenAI’s product.
This report pointed to the use of hashtags, including one referencing the far right, which OpenAI found to have been spread across platforms and posted alongside anti-Takaichi and anti-US memes. Other pieces of evidence uncovered aligned with the user’s plans as revealed to ChatGPT in October.
According to OpenAI, the ChatGPT user recorded that of 50,000 posts made across over 200 Western platforms, fewer than 150 posts received more than 300 shares or comments.
“None of the social media posts we identified had a significant degree of engagement,” OpenAI added. “While this activity does illustrate apparent operational planning and implementation across the internet, it does not appear to have achieved much impact.”
On Monday, Japanese daily Yomiuri Shimbun reported on Takaichi and her government being targeted by a coordinated, long-term campaign featuring thousands of social media accounts.
Citing research by the Tokyo-based Japan Nexus Intelligence analyst group, the report noted that the accounts were spreading fabrications in both Japanese and English, and using AI-generated images.
The false narratives, posted in late January ahead of election campaigning, included claims that Takaichi was “opening the path to military expansionism and historical revisionism”, as well as Japan’s supposedly increasing social security burden on the younger generation.
Apart from targeting Takaichi, the user also asked ChatGPT to help refine other status reports on “cyber special operations”, which OpenAI then found to have matched activity detected across the internet.
“The available evidence suggests that Chinese law enforcement is implementing a strategy … to suppress dissent and silence critics both online and offline, at home and abroad,” said OpenAI.
This resource-intensive and sustained effort relies on a “playbook of dozens of tactics” ranging from trolling and mass posting to impersonating US officials and targeting dissidents’ families.
OpenAI said it also identified an X account with behaviour connecting to both the anti-Takaichi operation and attacks on Chinese dissidents and human rights groups.
The Chinese law enforcement individual documented on ChatGPT that the “special operations” extended across Chinese social media platforms such as Weibo and WeChat as well as over 300 “foreign” ones. The user cited millions of posts on Chinese networks and tens of thousands on foreign ones.
“The user’s updates referred to a systematic use of AI for monitoring, profiling, translation, content creation and internal documentation,” said OpenAI.
Apart from the alleged Chinese influence ops, OpenAI’s threat report also covered steps taken to ban accounts using AI to pose as a fake dating agency to target young Indonesian men.
This activity “very likely originated in Cambodia and aligns with recent public reporting on Chinese-led criminal scam operations in the country”, the firm added.
“In isolated cases, individual users self-identified as scam workers in Cambodia, such as when asking the model for tax advice and stating their occupation as ‘scammer’.”
Get our pick of top stories and thought-provoking articles in your inbox
Stay updated with notifications for breaking news and our best stories
Get WhatsApp alerts
Join our channel for the top reads for the day on your preferred chat app
Copyright© Mediacorp 2026. Mediacorp Pte Ltd. All rights reserved.
We know it’s a hassle to switch browsers but we want your experience with CNA to be fast, secure and the best it can possibly be.
To continue, upgrade to a supported browser or, for the finest experience, download the mobile app.
Upgraded but still having issues? Contact us
