ChatGPT Cybercrime Discussions Spike to Nearly 3K Posts on Dark Web

Kaspersky's Digital Footprint Intelligence observed an uptick in Dark Web discussions (~3,000 posts) about illegal uses of ChatGPT and other LLMs, including the development of FraudGPT/malicious chatbots and the sale of roughly 3,000 stolen ChatGPT accounts; threat actors are sharing prompts and techniques to unlock and abuse AI capabilities. The report highlights growing misuse trends and recommends implementing reliable endpoint security to mitigate potential attacks.