Like a JavaScript engine, if it's written in C and C++, is not going to be safe until it's rewritten in a memory safe language.

But I'd say it's not all about the vulnpocalypse, right? I think, you know, obviously because of, like, mythos begat fable, and so we're talking about O'Day and exploit generation and whatnot.

Like, something else that happened here is that Security teams are builders, right? And security isn't just vulnerabilities and patching and whatnot, although all that stuff is very important.

There's all this other really boring toil that happens inside of security programs around like, asset inventory management and attack surface management and around on-call and around monitoring and triage and answering your security at email alias.

And there's lots of unglamorous stuff that people are working very hard to leverage generative AI to reduce that toil, make our programs faster, make them more responsive, handle an increasing estate of things that we have to secure.

And so when a defender like myself gets next generation capabilities like torn out of my hands a week after I started evaluating all the ways that this will increase the efficacy of my program and increase the efficiency of my program, that's not helpful.

That's not helping anyone.

I mean, that's, that's always, I think, been a good reliability, uh, uh, you know, for, for other nonpolitical reasons, but it's unfortunate that now politics, uh, is a potential, uh, uh, reliability and business continuity problem for AI.