July 12, 2026
- "Ghostcommit" hides prompt-injection payloads inside PNG images to fool AI coding agents. Researchers showed the technique slipping past AI code reviewers CodeRabbit and Bugbot — which never open image files — then convincing a coding agent to read a repo's
.envand write every secret into code encoded as a list of numbers. BleepingComputer · AI & Model Security
in Exploit Chains, Poisoned Packages, and AI Agents Turned Against Their Owners