New GhostGPT AI Chatbot Facilitates Malware Creation and Phishing

Researchers observed GhostGPT, a malicious generative AI chatbot sold via Telegram since late 2024 that claims to connect to jailbroken ChatGPT or open-source LLMs to provide uncensored responses. Marketed for coding, malware creation, exploit development, and phishing/BEC email generation, the tool enables low-skilled actors to rapidly produce convincing malicious content (researchers demonstrated a DocuSign phishing template) and has drawn thousands of views on criminal forums.