By
Katherine Li
Every time Katherine publishes a story, you’ll get an alert straight to your inbox!
By clicking “Sign up”, you agree to receive emails from Business Insider. In addition, you accept Insider’s
Terms of Service and
Privacy Policy.
Follow Katherine Li
- OpenAI published a report detailing China's "cyber special operations" that target critics online.
- The report followed OpenAI's findings and removal of a ChatGPT account belonging to the CCP.
- Bluesky said it recently removed a small number of accounts for "inauthentic coordinated activity."
It turns out that even the Chinese government uses ChatGPT — at least according to a cybersecurity report from OpenAI.
This is how OpenAI uncovered China's "cyber special operations" targeting political dissidents in China and abroad, after the company found and removed an account owned by the Chinese government that uploaded periodic status reports to ChatGPT and asked them to be polished.
In OpenAI's "Disrupting Malicious Uses of AI" report published on Wednesday, a third of the document detailed China's tactics to silence and intimidate critics across social media platforms owned by American tech giants.
"This effort appears to be large-scale, resource-intensive and sustained, counting at least hundreds of staff, thousands of fake accounts across scores of platforms, the use of locally deployed AI models, and a playbook of dozens of tactics," OpenAI wrote.
"The targets are not just people in China, but also dissidents around the world and representatives of foreign countries, up to and including the prime minister of Japan," OpenAI added.
OpenAI's report documented instances in which operatives for China forged US county court documents and submitted them to social media platforms in an effort to have certain posts removed.
Online operatives with thousands of fake accounts also collectively submitted abusive reports on dissidents to social media platforms to intentionally trigger false account bans and content restrictions, according to OpenAI. Many of these false reports, OpenAI said, include AI-generated images posing as screenshots of conversations and comments.
Some of the largest accounts on X have been targeted, according to OpenAI. That includes the handle @whyyoutouzhele, better known as "Teacher Li is not your teacher," which has the fifth-largest overall traffic share on X as of January, with more than 2.1 million followers. The account mainly posts videos from inside China, often documenting instances of corruption and human rights abuses.
"We hope social platforms like X, YouTube, and Bluesky realize that your content moderation system is being used by the Chinese Communist Party as a weapon," the team behind the Teacher Li account wrote in a post on X on Wednesday.
"We also call on the entire AI industry to take this matter seriously," the team added in the original post written in Chinese. "When your technology is being used to systematically oppress human rights, to say that 'we're just makers of a tool' is not an acceptable answer."
Other social media platforms have observed similar patterns and have taken action to curb these activities.
Aaron Rodericks, head of trust and safety at Bluesky, told Business Insider that the platform has invested in its capabilities to identify and disrupt influence operations like those described in the OpenAI report, including hiring specialized investigative staff and expanding monitoring systems.
"As part of this ongoing work, the team recently removed a small number of accounts, consistent with those described in the report, that were engaged in inauthentic coordinated activity in violation of our Community Guidelines," said Rodericks.
A person familiar with the matter told Business Insider that Meta includes activities that OpenAI described in its regular adversarial report, and takes action against accounts found to be violating platform policies.
X and China's Ministry of Foreign Affairs did not immediately respond to requests for comment.
