Post

“Next Exploit Opportunities”: How a Threat Actor Used AI to Help Plan and Execute Multiple Types of Cybercrime Activities

Unit 42 documents a threat actor (BUGSTER) weaponizing the Hermes Agent framework with DeepSeek V4 and Cursor IDE to orchestrate multiple cybercrime campaigns targeting Indian universities and organizations. The actor used AI-generated reconnaissance scripts and exploitation tools to conduct WordPress brute force attacks (extracting 13 usernames via REST API enumeration), exfiltrate university databases (5,200+ student records), plan VoIP telecommunications fraud, crack Wi-Fi passwords, and establish C2 operations. The analysis reveals a concerning trend: attackers lacking traditional hacking skills are using readily-jailbroken AI models (DeepSeek’s low safety friction and cost-effective pricing) to bridge capability gaps, suggesting Hermes + DeepSeek is becoming a prevalent threat actor toolchain with broad impact across multiple attack vectors.

Read original article

This post is licensed under CC BY 4.0 by the author.