Partager:

Microsoft researchers announced on Thursday 7 September that they had discovered a network of fake social media accounts they believed were controlled by the Chinese side. The network aims to influence United States voters by using artificial intelligence.

A spokesman for the Chinese embassy in Washington said China's accusations of using artificial intelligence to make fake social media accounts full of prejudice and malicious speculation and that China is pushing for the use of safe artificial intelligence.

In a new research report, Microsoft said the social media account was part of China's information operation. According to Microsoft, the campaign has similarities with activities that the United States Department of Justice attributes to elite groups within China's Ministry of Public Security.

Researchers did not mention which social media platforms were affected, but screenshots in their reports showed shipments from what it appears to be Facebook and Twitter, now known as X.

The report underscores the tense social media environment as the United States prepares for the 2024 presidential election.

The US government has accused Russia of interfering in the 2016 elections with a hidden social media campaign and has warned of further efforts by China, Russia, and Iran to influence voters.

The report provides limited examples of recent activity and does not explain in detail how researchers attributed the shipment to China.

A Microsoft spokesman told Reuters that company researchers used a "multiplex distribution model," which relies on "technical evidence, evidence of behavior, and contextual evidence."

"The campaign is starting to use genative artificial intelligence technology around March 2023 to create content that is given political content in English and "mimitting US voters," said Microsoft spokesman.

Immunity of artificial intelligence can create images, text, and other media from scratch.

This new content is much more "binding the eyes than the awkward visuals used in previous campaigns by Chinese country players, who rely on digital images, stock photo collages, and other manual graphic design," the researchers wrote.

The report mentions an example of one image produced by artificial intelligence, which Microsoft says comes from a Chinese account, which depicts Statue Liberty holding an assault rifle with the caption: "Everything is being thrown away. DEWI VIOLENCE."

A Microsoft spokesman said identified accounts had tried to look American by listing their public locations as within the United States, posting American political slogans, and sharing hashtags related to domestic political issues.


The English, Chinese, Japanese, Arabic, and French versions are automatically generated by the AI. So there may still be inaccuracies in translating, please always see Indonesian as our main language. (system supported by DigitalSiber.id)