The shift to AI-native development has moved the AppSec bottleneck from generating code to securing it, but the true threat isn’t just the sheer volume of new software—it’s the autonomous agents working behind the scenes. Broadcasting live from RSAC, Techstrong Group’s Alan Shimel sits down with Backslash Security’s Gil Friedman to discuss how malicious prompt injections and compromised APIs can manipulate the “AI factory” long before a single line of code is even committed. Friedman explains why traditional security tools are completely blind to these new attack surfaces and how organizations must establish visibility and strict guardrails to protect their data from non-deterministic agentic workflows.