OpenAI Takes Action Against State-Linked Threat Actors Abusing ChatGPT

OpenAI has recently blocked several ChatGPT accounts linked to state-sponsored and cybercriminal groups from countries like Russia, China, and Iran for malicious AI usage. A notable case involved a Russian-speaking group that developed malware called Crosshair X using ChatGPT, disguising it as legitimate software to steal sensitive data. The group’s operational security practices were sophisticated, employing tactics like DLL side-loading and IP obfuscation. OpenAI also detected activity from Chinese APT groups using ChatGPT for research and automated social media manipulation, and an Iranian operation crafting politically charged content. In total, nine threat groups were identified misusing ChatGPT for activities such as creating fake content and organizing fraudulent schemes. OpenAI emphasized the dual-use nature of its models and announced enhanced user activity monitoring and collaborations to combat these harmful applications, focusing on both cyber threats and manipulative geopolitical information campaigns.

Source link