
A bug-bounty test shows how AI tools can accelerate vulnerability discovery and raise new security questions for AI labs.
OpenAI banned a cluster of ChatGPT accounts tied to a covert Russian influence campaign that generated social posts, promoted a fictitious think tank, and targeted Western audiences across multiple platforms.

OpenAI disrupted a covert Russian influence campaign by banning a cluster of ChatGPT accounts, according to The Decoder’s report. The operators accessed the platform from Russia through VPNs and asked ChatGPT to remove linguistic clues that could point to Russian origin.
The campaign used AI-generated social media posts to push pro-Kremlin narratives across Western platforms. The key reader takeaway: AI tools can lower the effort needed to produce influence content, even when the initial audience is limited.
The campaign promoted the “International Burke Institute,” described as a think tank supposedly based in Israel. Its content included a “sovereignty index” that ranked Russia favorably while making Western countries look worse by comparison.
OpenAI said 34 of 36 expert-linked IBI articles published between September 2025 and May 2026 were copied from other sources, with some using fake author credits. The report cites examples involving Cambridge University Press, the University of Nottingham, the Migration Policy Institute, and an Australian food chemistry professor.
The campaign spread posts across X, LinkedIn, Facebook, Substack, and Telegram. Some content appeared in German on a Telegram channel called “Lahme Ente,” which criticized Ukraine, the EU, and Germany’s federal government while pushing for closer ties with Russia.
OpenAI said individual posts had very few views and official IBI accounts had low subscriber counts. However, linked Telegram channels each reached 10,000 to 20,000 followers, and OpenAI rated the campaign category three out of six on the Brookings Breakout Scale.
OpenAI said the campaign’s current reach stayed small, but the infrastructure behind it could have been scaled up over time. A second operator reportedly created logos for about a dozen Telegram channels focused on Germany, the US, France, Poland, and Turkey, and requested Russian-language summaries of channel activity.
The report also notes that OpenAI has previously disrupted Russian influence operations abusing ChatGPT, including “Bad Grammar” and “Operation Helgoland Bite.” The practical takeaway is to evaluate political content by checking source identity, authorship, cross-platform repetition, and whether a polished institution appears to have verifiable credibility.

A bug-bounty test shows how AI tools can accelerate vulnerability discovery and raise new security questions for AI labs.

OpenAI is reportedly close to another major math push, but any announcement may wait.
A small AI cybersecurity startup used Claude in research that exposed OpenAI account vulnerabilities.

OpenAI’s new reporting framework starts with six misalignment cases, including self-generated prompt injections.