China's Secret Influence Strategy Accidentally Revealed Through ChatGPT Use
A major covert operation by the Chinese government has been inadvertently exposed after a law enforcement official attempted to use OpenAI's ChatGPT artificial intelligence tool to plan a sophisticated smear campaign. The incident, detailed in a report by OpenAI, reveals systematic strategies for transnational repression and influence operations targeting critics of the Chinese Communist Party (CCP) abroad.
ChatGPT Query Exposes Elaborate Smear Campaign Plans
The Chinese official specifically asked ChatGPT to develop a multi-part plan to discredit then-incoming Japanese Prime Minister Sanae Takaichi. The proposed strategy included fanning online anger about United States tariffs on Japanese goods to create political pressure. While ChatGPT reportedly did not respond to this particular prompt, the timing of subsequent events proved revealing.
After Takaichi assumed power, coordinated hashtags emerged on a popular forum for Japanese graphic artists attacking her leadership and complaining about US tariffs. This real-world activity matched the strategy the official had described to ChatGPT, suggesting the influence operation proceeded despite the AI's non-response.
Industrialized Transnational Repression Operations
Ben Nimmo, principal investigator at OpenAI, characterized the Chinese operation as "industrialized" in scale and methodology. "It's not just digital. It's not just about trolling. It's industrialized," Nimmo explained. "It's about trying to hit critics of the CCP with everything, everywhere, all at once."
The investigation revealed that Chinese operators had been impersonating United States immigration officials to intimidate Chinese nationals living in America. They contacted a US-based Chinese dissident while pretending to be immigration authorities, attempting to create fear and compliance through deception.
Forged Documents and False Obituaries
OpenAI's report detailed additional tactics employed by the Chinese operation. Investigators discovered attempts to forge documents purportedly from a US county court to pressure social media platforms into taking down a Chinese dissident's account. The official's ChatGPT conversations also described creating a false obituary for a Chinese dissident to spread damaging rumors.
This specific tactic matched real-world events from 2023, when a fabricated obituary targeting a Chinese critic actually circulated online. OpenAI investigators didn't rely solely on the ChatGPT conversations but cross-referenced the official's descriptions with actual online activity, finding multiple points of correlation that validated their findings.
Account Ban and Government Silence
OpenAI responded to the discovery by banning the Chinese official's account after identifying what they described as China's "secret strategy for covert influence operations and transnational repression." The company's investigation revealed that the official had been using ChatGPT to track and manage their covert network operations.
The Chinese Embassy has not commented on OpenAI's report or the allegations contained within it. The silence from official Chinese channels contrasts with the detailed evidence presented in the investigation, which paints a picture of systematic, government-backed operations targeting political opponents internationally.
Broader Implications for AI and State-Sponsored Influence
This incident highlights the emerging challenges at the intersection of artificial intelligence and state-sponsored influence operations. While ChatGPT itself didn't assist with the smear campaign planning, its use as a tool for organizing such operations reveals how AI platforms might be co-opted by bad actors.
The exposure of these tactics through an AI platform conversation represents a significant intelligence failure for the Chinese operation, demonstrating how digital tools can inadvertently reveal secret government strategies. As AI becomes more integrated into daily operations across sectors, similar incidents may become more common, creating new vulnerabilities for covert operations while providing investigators with unprecedented windows into previously hidden activities.
