So I guess, does that concern you at all with how fast the threat landscape is growing, how cheap it is to perform these attacks, and how easy it is to scale them? So

on the final point you made, so I agree with that.

So Google's developer operations research analysis says This thing called DORA, not to be confused with the European Digital Operational Resiliency Act.

So Google's DORA team did a lot of research on this.

And the conclusion is kind of obvious in hindsight, but it's good for it to be founded on research that if you take AI-driven software production into an organization that's got quite unmanaged and chaotic software development life cycles, you're going to get chaos amplified.

If you deploy...

AI driven or agentic driven software production into an organization that's got quite a well controlled software production pipeline and controlled build and testing processes, you get productivity amplified.

So when we think about the actor within our organization as an agent, what do you think are kind of the new ways we need to think about that trust layer? And maybe how we need to start reasoning about trust for AI systems that are working autonomously within our environments.