"Combatting State-Linked Cyber Threats with OpenAI and Microsoft"

OpenAI has shut down accounts from China, Iran, Russia, North Korea, and Iran that were allegedly used for malicious activities such as generating phishing emails and evading malware detection. The accounts were reportedly using OpenAI's models for harmful intent, prompting the shutdown in collaboration with Microsoft. OpenAI emphasized that its models have limited capabilities for malicious cybersecurity tasks and aims to prevent misuse. Microsoft's analysis revealed specific activities of the state-affiliated actors, including researching information, evading malware detection, and crafting phishing attacks.