Suspension of Over 13,000 Accounts Suspected to be Involved in Influence Operations in China, Russia, and Iran

Social media platform X announced on Wednesday, October 14th, that it has shut down more than 13,000 accounts related to suspected foreign influence operations involving Iran, Russia, and China. These accounts were found to be using fake identities to spread specific content with the intention of intervening in American political discussions.

With the upcoming midterm elections in the United States on November 3rd, X not only disclosed its account suspension actions but also introduced measures to prevent false accounts and misleading information during the election period.

The global government affairs team at X stated, “Whether it’s networks spreading disinformation or organizations engaging in influence operations, we are actively identifying and stopping such organized false activities to safeguard platform credibility.”

X emphasized that their security measures operate 24/7 year-round, with the safety team using various detection methods during the election period to proactively identify and address any violations. The “Authenticity Policy” is a crucial foundation for combating false accounts and organized false operations.

Additionally, through the “Civic Integrity Policy,” X addresses misleading information that could disrupt public participation in elections, such as confusing voters on when, where, or how to vote, or attempting to prevent voter engagement through posts falling under this policy’s scope.

Users are encouraged to report suspicious content to the platform. These reports may lead to requests for adding “Community Notes,” allowing other users to provide background explanations and reliable information to help readers understand the facts.

For significant violations that require further action, X may directly add warning labels to posts, indicating the content violates the platform’s election-related regulations.

X confirmed that these protective measures will continue beyond election day.

Prior to this, the AI company OpenAI released an investigation report on October 8th, revealing the disruption of influence operations originating from Russia and Iran, with related ChatGPT accounts being banned.

The operators used ChatGPT in conjunction with traditional methods to create false journalist identities and organizations, concealing the sources of information to spread geopolitics and conflict-related content.

In the case of Iran, the operation network utilized 7 fictional journalist identities, embellishing articles and submission letters with ChatGPT, and generating social media comments.

OpenAI found nearly 100 articles published or republished under these pseudonyms in over ten global online media outlets, although the social media interactions generated by these operators were generally minimal.

On the Russian front, individuals utilized ChatGPT to write internal reports, produce propaganda content, and seemingly control a think tank named the “Social Research Center” (SRC) through fictional identities.

The investigation revealed that they employed locals in Latin America for research work, seemingly unaware that they were actually working for Russian entities. Moreover, they fabricated “leaked documents” and used ChatGPT to write voice-over scripts for fake news videos. Some of this content has circulated online, as discovered by OpenAI.

OpenAI emphasized, “AI enables these camouflage operations to have larger scale, greater efficiency, more fluent language expression, and stronger editing capabilities.”

Iran International, based in London, noted that X’s predecessor Twitter had previously discovered Iran-related influence networks.

In 2021, Twitter suspended 238 accounts operated from Iran. The investigation began when some accounts attempted to interfere in the public discourse during the first debate of the 2020 U.S. presidential election. Twitter believed these accounts had low engagement levels that had not yet influenced public opinion.

In 2020, Twitter also suspended 104 accounts related to Iranian networks for artificially amplifying sensitive U.S. topics, including the Black Lives Matter movement, the killing of George Floyd, and other racial issues.