Cybersecurity voices push back: defenses can contain AI even if it's great at zero-days

basedjensen · x · 2026-09-27

Zack Korman called on prominent cybersecurity figures to state the obvious: cybersecurity is real and works, and even an AI that's extremely good at finding zero-days can be contained. Security practitioner @0xTib3rius amplified the point, adding that there are a finite number of zero-days and that AI itself can be used to iteratively improve sandbox security. The thread pushes back on the fear that offensive AI capability will inherently overwhelm defenses.

Related event: Security Experts Say Cyber Defense Holds Even Against AI-Driven Zero-Day Hunting(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →