Google Finds Attackers Using Autonomous AI Agents to Turn Security Patches into Exploit Code

Written by

in

Google and Mandiant have discovered that attackers are using autonomous AI agents to reverse-engineer security patches, generate exploit code, and attempt to execute it without human approval. Researchers found a multi-agent framework named exploit_agent on a publicly exposed server, along with scripts for reconnaissance, brute-forcing, and installing webshells; similar frameworks have also been used in intrusions targeting American aerospace organizations.

Source:The New Stack · Read Original

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *