Caught in the Act: 'Simplified Chinese' Characters Reveal Attempted Manipulation

OpenAI has claimed that it has detected signs of an attempt to manipulate public opinion in the United States using ChatGPT by forces suspected to be linked to China.

In a 'Threat Report' released recently, OpenAI stated that it had identified and blocked two clusters of ChatGPT accounts suspected of being linked to China.

It was revealed that these groups were active from late last year through early this year, attempting to sow division by targeting current issues such as energy concerns related to artificial intelligence (AI) policy and trade policies, including tariffs, within the U.S.

One group led a campaign claiming that AI data centers were driving up electricity bills for ordinary households.

Although they used VPNs to disguise their location as being outside of China, the prompts entered into ChatGPT were in Simplified Chinese, which is used in mainland China.

The report analyzed that these individuals are likely part of a private Chinese company that counts local Chinese governments as its clients.

Another group launched a campaign criticizing U.S. tariff policies as an attempt to seize technological hegemony.

They used ChatGPT to generate a large volume of political cartoons on the theme of the U.S.-China tech competition, and it was found that they consistently included instructions such as, "The cartoon should only feature President Donald Trump and must not feature Chinese President Xi Jinping."

The cartoons generated in this manner were distributed through social media platforms such as X (formerly Twitter).

The report added that these individuals referred to themselves as 'Shuijun,' which is Chinese slang for online accounts that engage in organized criticism.

However, OpenAI assessed that most of the attempts at public opinion manipulation by these groups failed to exert any real influence.

In response, the Chinese Embassy in the U.S. refuted the claims, stating, "We firmly oppose groundless attacks and slander against China," and added, "The Chinese government is committed to ensuring that AI serves as a force for good for everyone."

Reported by Kim Taewon | Video by Lee Dain | Graphics by Yook Dohyun | Produced by SBS Digital News
※ Please note: This article was translated by AI and may contain errors.