Notes

This article talks about the development of malicious AI chat bots being advertised and developed by a user who goes by CanadianKingpin12. The first is a malicious version of ChatGPT, known as FraudGPT, which came out on July 25, 2023. A few things FraudGPT can do is write malicious code, create undetectable malware, and create phishing pages. After advertising FraudGPT, a cybersecurity company by the name of SlashNext conducted an investigation on CanadianKingpin12. They discovered that he has been developing a new malicious AI that is based off of BARDs language model. In private conversations, CanadianKingpin12 has said that "they were working on DarkBART - a "dark version" of Google's conversational generative artificial intelligence chatbot" (Toulas). In a video posted online, we can see that this new malicious AI will be able to develop sophisticated phishing campaigns to deceive unsuspecting victims and steal their credit card information, tell you how to execute advanced social engineering attacks, tell you how to exploit vulnerabilities in computer systems and networks to gain unauthorized access, perform data breaches, or sabotage critical infrastructure, tell you how to create and distribute malicious software, and tell you about zero-day vulnerabilities and how to exploit them for financial gain. This is just what the user revealed in a video, so we can assume the AI will be able to do much more on top of this.

Classification

This malware can easily be classified as Adversarial Artificial Intelligence. Any person in possession of this chat bot can become a competent hacker as all they have to do is follow instructions.