OpenAI Bans Russian-Linked Accounts Promoting Fabricated Israeli Think Tank

This incident underscores the ongoing battle against misinformation and the lengths to which state-linked actors will go to manipulate public perception.

• OpenAI banned a cluster of ChatGPT accounts on August 25, 2026, linked to a Russian influence operation.
• The operation promoted a fabricated think tank, the International Burke Institute (IBI), which disseminated pro-Kremlin narratives.
• The campaign reached limited audiences primarily through Telegram channels, amid broader Russian efforts to sway Western opinions during the Ukraine conflict.

• Russian influence operations have been active since the 2022 invasion of Ukraine, utilizing AI to amplify their reach.
• The IBI website, registered in February 2025, published content that misattributed academic articles to lend credibility to its narratives.
• OpenAI's previous interventions against similar campaigns highlight the escalating sophistication of disinformation tactics.

On August 25, 2026, OpenAI took decisive action against a cluster of ChatGPT accounts that were part of a covert influence operation linked to Russian state actors. This operation centered around the International Burke Institute (IBI), a fabricated think tank that aimed to promote pro-Kremlin narratives by generating misleading social media content. The IBI's website, which began publishing in September 2025, featured a mix of copied academic articles and a sovereignty index that ranked Russia favorably while downgrading Western nations.

The investigation into this operation began with the detection of AI-generated posts that were disseminated across various social media platforms, including Telegram, Substack, X, Facebook, and LinkedIn. Operators cleverly bypassed OpenAI's access restrictions by using VPNs and prompting the model in Russian, instructing it to obscure its linguistic origins. This allowed them to create promotional content that appeared legitimate and credible.

Of the 36 articles sampled from the IBI website, a staggering 34 were found to be copied from legitimate academic sources, often with incorrect attributions to deceased scholars or unrelated experts. This tactic of misattribution was designed to lend an air of authenticity to the IBI's claims, making it easier to sway public opinion in favor of Russian narratives.

Despite the operation's elaborate facade, audience engagement was relatively limited, with most posts reaching Telegram channels that had between 10,000 and 20,000 followers. OpenAI's decision to ban the accounts was a significant step in combating the misuse of AI for disinformation, but the IBI website remains online, casting doubt on its legitimacy.

The IBI has denied any coordination with Russian state actors, but the evidence suggests a well-orchestrated effort to influence perceptions during a time of heightened geopolitical tension. This incident highlights the ongoing challenges faced by tech companies in monitoring and mitigating the spread of disinformation, particularly as state-linked actors continue to evolve their tactics.

• Social media users: Increased exposure to misleading information can skew perceptions and opinions.
• Academics and researchers: Misattributed articles undermine the credibility of legitimate scholarship.
• Policy makers: Disinformation can complicate decision-making processes and public trust in governance.

• Increased scrutiny on AI tools: Expect more regulations and oversight on how AI can be used in content generation to prevent misuse.
• Emergence of new disinformation tactics: As tech companies adapt, state-linked actors may develop more sophisticated methods to bypass detection.
• Public awareness campaigns: Look for initiatives aimed at educating users about identifying and combating misinformation online.