Uploaded May 2026 | Updated September 2026, 23 hours ago

---

Everyone's talking about LLMs finding and exploiting security vulnerabilities. I wanted to see how real the threat actually is, so I tried the dumbest possible prompt on my own code. And it found a significant vuln.

In this video, Adrian walks through what he did and what he found, shows you how to run the same kind of analysis on your own projects, and closes with practical advice on securing your systems before the attackers get there first.

Links:

Nicholas Carlini "Black-hat LLMs" youtube.com/watch?v=1sd26pWhfmg
Cloudflare Blog "Project Glasswing: what Mythos showed us" blog.cloudflare.com/cyber-frontier-models
Chainguard Containers https://www.chainguard.dev/containers
Chainguard Libraries https://www.chainguard.dev/libraries
Chainguard Actions https://www.chainguard.dev/actions

About Chainguard
Founded by the industry's leading experts on open source software, security and cloud native development, we are on a mission to be the safe source for open source.

Where to find us:
Website: https://www.chainguard.dev/
Twitter: twitter.com/chainguard_dev
LinkedIn: linkedin.com/company/chainguard-dev
TikTok: tiktok.com/@chainguard_dev