OpenAI has suspended a coordinated cluster of ChatGPT accounts linked to a Russian-backed influence operation that used generative AI to produce pseudo-academic research and a fabricated geopolitical index. The takedown underscores a growing trend in state-aligned information campaigns, where commercial large language models are increasingly deployed to manufacture institutional credibility at scale.
According to reporting from Security Affairs, the suspended accounts were used to promote a fictitious organization identified as the IBI. The operation distributed AI-generated articles alongside a manufactured “sovereignty” index designed to advance pro-Russian narratives while circumventing traditional editorial oversight. Rather than relying on manual content farms or basic bot networks, the operators leveraged commercially available AI platforms to rapidly draft high-volume, academically styled material. By mimicking the citation structures, analytical framing, and tone of legitimate research, the campaign aimed to bypass automated moderation and gain traction among policymakers, journalists, and academic circles.
OpenAI confirmed the suspension of the account cluster, noting that the activity likely originated in Russia and was part of a covert influence operation. The enforcement action highlights the ongoing challenge of distinguishing AI-generated disinformation from organic platform traffic. While centralized AI providers can monitor usage patterns and enforce terms of service, the rapid evolution of generative tools continues to complicate detection and attribution efforts for platform operators and security researchers.
OpenAI has not disclosed the exact number of suspended accounts or the full distribution footprint of the IBI-generated material across social and professional networks. As generative AI becomes more accessible, cybersecurity teams and platform operators are increasingly focused on developing standardized detection methods and cross-platform intelligence-sharing protocols to identify and disrupt coordinated influence campaigns before they achieve widespread reach.
OpenAI 已停用一組協調運作的 ChatGPT 帳戶，該批帳戶牽涉一項獲俄羅斯支持的影響力行動，有關行動利用生成式 AI 製造偽學術研究及捏造地緣政治指數。此次取締突顯國家級資訊行動日漸成勢，營運者日益部署商用大型語言模型，以規模化方式營造機構公信力。
據 Security Affairs 報道，遭停用的帳戶被用於推廣一個名為 IBI 的虛構組織。該行動發布 AI 生成的文章及捏造的「主權」指數，旨在推進親俄敘事，並刻意繞過傳統編輯把關。營運者不再依賴人手操作的內容農場或基礎機械人網絡，轉而利用商用 AI 平台，快速撰寫大量具學術風格的素材。透過模仿正規研究的引用結構、分析框架及語調，該行動旨在繞過自動審核，並在政策制定者、記者及學術界中取得關注。
OpenAI 已確認停用該批帳戶，並指出有關活動很可能源自俄羅斯，屬一項隱蔽的影響力行動。此次執行措施突顯區分 AI 生成的虛假資訊與平台自然流量仍面臨持續挑戰。儘管集中式 AI 供應商能夠監控使用模式及執行服務條款，但生成式工具的快速演進，持續令平台營運商與安全研究人員在偵測及歸因工作上更趨複雜。
OpenAI 尚未披露遭停用帳戶的確切數目，以及 IBI 生成素材在社交與專業網絡中的完整傳播範圍。隨著生成式 AI 日益普及，網絡安全團隊及平台營運商正日益專注於開發標準化偵測方法與跨平台情報共享協議，以期在協調影響力行動廣泛傳播前予以識別及瓦解。