OpenAI banned a network of ChatGPT accounts linked to Russia that used the tool to run a covert influence campaign built around a fake academic think tank called the International Burke Institute. The operation included a website with copied and misattributed academic work, a sovereignty index portraying Russia favorably, listed affiliations with entities such as the UN, UNICEF, and named individuals including Mike Pompeo and Francis Fukuyama, and efforts to hide the operators' Russian origins, with content spread across Telegram channels and other platforms. OpenAI did not attribute the campaign to the Russian government, and reports noted the posts generally received limited views despite some associated Telegram channels having up to 20,000 followers.
Le Monde frames the story with more narrative and skeptical detail, highlighting the deceptive construction of a fake Israeli-sounding think tank with prestigious-seeming (but fabricated) affiliations, underscoring how convincingly the disinformation was disguised across multiple languages and platforms.
CNBC frames the story straightforwardly as OpenAI banning a Russian covert influence campaign, emphasizing the 'sovereignty index' and comparing it to the earlier Rybar ban, presenting OpenAI's warning that AI can help manufacture authority while also exposing operations.
Decrypt frames the story analytically and cautiously, stressing that reach was limited, that OpenAI did not attribute the campaign to the Russian government, and situating it within a broader pattern of past Russia-linked AI misuse (Doppelganger, Forest Blizzard) while noting unresolved questions about Israeli individuals' involvement.