Russia Used ChatGPT to Build Propaganda Network

Russian operators used ChatGPT to conduct a covert influence campaign aimed at promoting pro-Russian narratives and creating the perception of an independent expert community.

OpenAI reported this after identifying and blocking several ChatGPT accounts that likely originated in Russia.

The operators used VPNs to access the service because OpenAI does not provide access to its models from Russia.

The main target of the campaign was the Burke Institute (IBI), a self-proclaimed “expert community” that presented itself as an organization from Israel.

Using ChatGPT, the operators created posts and comments to promote IBI materials on X, LinkedIn, Facebook, and Telegram. Most of the content was generated in English, while the models were also instructed to conceal any signs of the authors’ Russian origins.

In addition to posts directly related to IBI, the operators created German-language content for a Telegram channel. It regularly criticized Ukraine, the EU, and the German government while promoting the idea of improved relations with Russia.

One of the operation’s participants also used ChatGPT to create logos for about a dozen Telegram channels aimed at audiences in Germany, the United States, France, Poland, and Turkey.

The operation was centered on the IBI website, which was registered in February 2025. It presented itself as an independent expert organization and featured material attributed to prominent academics and researchers.

However, a review of 36 articles published on the website between September 2025 and May 2026 found that 34 were copied from other online sources.

Another element of the campaign was the so-called “Sovereignty Index.” IBI used it to compare different countries while portraying Russia in a favorable light and criticizing states that condemned Russia’s invasion of Ukraine.

OpenAI assessed the operation’s direct impact as relatively limited. Most posts received few views, while IBI’s official accounts had few followers. Meanwhile, some Telegram channels in the network gained between 10,000 and 20,000 followers.

OpenAI noted that the main danger posed by the operation was building infrastructure for future propaganda distribution. AI was used as an auxiliary tool to scale up content production, conceal its origin and create the appearance of an authoritative organization.

Earlier this year, it was reported that Russia was planning covert interference in electoral processes in African countries, including Burundi, Cameroon, and Madagascar, using AI. The budgets for some projects in that network reached $600,000 per year.