AGI Dreams Podcast – March 19, 2026

AGI Dreams – Open, Uncensored, & Local - AI News Digest · Mar 19, 2026 · 17m · transcript

Risk Velocity — AI Code Security Hits the Compounding Phase

In this episode:

Risk Velocity — AI Code Security Hits the Compounding Phase — AI-generated code is exposing application security's inability to scale, with Veracode's Chris Wysopal introducing 'risk velocity' as the defining CISO metric. New tools like Greywall offer deny-by-default sandboxing for AI coding agents, while developers struggle with a comprehension gap where generated code looks correct but hides auth bypasses and dependency risks.

Making Software Agent-Native — CLI-Anything and the Gateway Wars &mdash…