In the latest report by the American artificial intelligence research company OpenAI, a large-scale overseas influence operation by the Chinese Communist Party (CCP) was revealed. CCP law enforcement officials attempted to disguise themselves as US immigration officials to intimidate Chinese dissidents living abroad.
This is one of the latest cases exposing CCP’s overseas transnational repression, showcasing how authoritarian regimes utilize artificial intelligence (AI) tools overseas to engage in activities that harm Western societies.
Released in February, the OpenAI report first disclosed CCP’s secret network operations, stemming from a CCP law enforcement official using ChatGPT as a diary to meticulously document these secret repression actions.
According to the ChatGPT user’s account, in one case, the CCP individual disguised as a US immigration official warned a Chinese dissident residing in the US, claiming their public statements were allegedly illegal.
In another case, CCP operatives used forged US county court documents to deceive a US social media platform into shutting down the social media account of a Chinese dissident. The CCP law enforcement official stated that although this action ultimately failed, it demonstrated the potential for such operations.
OpenAI highlighted that this influence operation involved hundreds of CCP operatives and thousands of fake accounts operated on various social media platforms.
“This is the face of modern transnational repression by China (CCP),” said Ben Nimmo, OpenAI’s Chief Researcher, before the report’s release. “It is not just about online repression, nor just online trolls. It is industrial scale, an attempt to use all means to simultaneously target critics of the CCP anywhere.”
According to analysts, the majority of content from this secret network was generated using various tools and then disseminated extensively through social media accounts and websites. OpenAI banned the user upon discovering their activities.
OpenAI investigators stated that this secret network also deliberately fabricated news about the deaths of Chinese dissidents, including creating fake obituaries and tombstone photos, and publishing them online.
In another instance, a ChatGPT user requested a multi-step plan aimed at defaming the incoming Japanese Prime Minister, Naomie Takai. One measure was stirring up public anger against the US by imposing tariffs on Japanese goods. OpenAI claimed that ChatGPT refused to respond to this directive. However, in late October 2025, as Naomie Takai assumed office, attacks against her and complaints about US tariffs appeared on a popular forum of Japanese manga artists.
At a time of escalating competition between the US and China for AI dominance, this report is crucial. The key to this game is how this technology will play a role in battlegrounds and economic decision-making.
Michael Horowitz, a former Pentagon official now a professor at the University of Pennsylvania, told CNN that OpenAI’s report “clearly illustrates how Beijing actively uses AI tools to enhance information warfare.”
The US House of Representatives’ special committee on China stated that they have noted OpenAI’s latest report and pledged to continue exposing CCP’s malicious actions targeting dissidents and undermining the integrity of US institutions.
On Monday, another American AI company named Anthropic called out three Chinese companies for illegally training AI using their Claude chatbot.
Anthropic also warned that foreign labs might forgo security precautions, integrating AI into military, intelligence, and surveillance systems. This could allow authoritarian governments to utilize advanced AI for cyber attacks, disinformation campaigns, and large-scale monitoring.