< Back to all clusters
[TECHNOLOGY] · Russia · 71 sources

started · updated

OpenAI dismantles Russian influence operation using ChatGPT

OpenAI has dismantled a covert influence operation originating from Russia that utilized ChatGPT to spread disinformation. The campaign focused on promoting the International Burke Institute (IBI), a fraudulent organization that posed as an Israeli-based expert community.

To maintain a veneer of credibility, the network used generative AI to produce social media content and articles. Many of these posts were based on academic works that had been copied from other sources and falsely attributed to real scholars. The operation also utilized a fabricated 'sovereignty index' designed to present Russia in a positive light while criticizing Western nations and institutions.

Operators used Russian-language prompts to generate content primarily in English, specifically instructing the AI to remove linguistic markers that might reveal their Russian origin. Because OpenAI services are unavailable in Russia, the actors used virtual private networks (VPNs) to access the platform. While OpenAI noted that the campaign reached relatively small audiences, it highlighted the sophisticated infrastructure built to scale such deceptive operations across platforms including X, LinkedIn, Facebook, Substack, and Telegram.

Entities

ChatGPT · Hugging Face · International Burke Institute · Israel · METR · Noam Chomsky · OpenAI · Redwood Research · Russia · Ukraine

Claims

What the coverage asserts, and how many sources carry each claim.

Sources

about 15 hours ago
about 14 hours ago
about 18 hours ago
about 20 hours ago
about 19 hours ago
about 16 hours ago
about 11 hours ago
about 21 hours ago
about 10 hours ago
about 13 hours ago
about 13 hours ago
about 24 hours ago