A new report from OpenAI uncovered a massive Chinese influence campaign — and it all came to light because one Chinese law enforcement official used ChatGPT like a personal diary.

According to OpenAI, this official documented how Chinese operatives were working to intimidate dissidents living abroad. Some of the tactics included pretending to be U.S. immigration officers to scare activists, forging fake U.S. court documents to shut down social media accounts, and even spreading false rumors about a dissident’s death with fake obituaries and gravestone photos.

The operation was huge: hundreds of people involved, thousands of fake accounts across social media, all aimed at silencing critics of the Chinese Communist Party.

Ben Nimmo, an investigator at OpenAI, described it as “industrialized repression” — not just online trolling, but a full-scale effort to hit critics “with everything, everywhere, all at once.”

OpenAI banned the ChatGPT user once they realized what was happening, but not before matching the diary entries to real-world online activity. For example, the official had asked ChatGPT to help plan a smear campaign against Japan’s incoming prime minister, Sanae Takaichi. While ChatGPT refused, similar attacks later appeared online.

The report highlights how authoritarian governments are experimenting with AI to boost their censorship and intimidation efforts. It also comes at a time when the U.S. and China are locked in a bigger struggle over who will dominate AI — not just in business, but in national security.

This story shows how AI isn’t just about innovation — it’s also becoming a tool in global power struggles.