OpenAI disrupted a likely China-origin operation, codenamed "Silver Lining Playbook", targeting US persons. Actors used ChatGPT to request information about US federal buildings, personnel distribution, and face-swapping software. The accounts generated emails resembling social engineering attempts, purporting to be from a Hong Kong-based consulting firm, Nimbus Hub Consulting. However, linguistic analysis of Simplified Chinese characters and activity during mainland business hours suggested operators were based in mainland China. The actors sought guidance on installing Face Fusion for live face-swapping and requested email drafts targeting state-level US officials. OpenAI banned a small set of ChatGPT accounts involved in the operation.

The actors used VPNs to access the platform and focused on gathering intelligence regarding US federal offices and online forums frequented by economics and finance professionals. The operation referenced Nimbus Hub Consulting, a Hong Kong-registered company, though the use of Simplified Chinese and operational patterns indicated mainland Chinese origins. OpenAI's disruption highlights ongoing efforts to counter malicious AI use for geopolitical influence and social engineering campaigns.