10 Out of 11 Coding Agents Failed. Here's Why That Number Should Concern You. Researchers...

Decades-old Bash shell tricks can bypass safeguards in most open source AI coding agents, creating a new software supply chain risk.

Adversa AI says GuardFall bypassed safety checks in 10 of 11 AI coding agents by exploiting how Bash rewrites commands before execution.

"Agentjacking" is the latest demonstration of how easily attackers can exploit an AI agent's inability to differentiate between content and instructions.

10 Out of 11 Coding Agents Failed. Here's Why That Number Should Concern You. Researchers...