Oh

And as a company certainly want to, um, manage it, but, uh, the employee usually, uh, say, "Okay, this is a good tool.

Maybe I just want to use it.

[laughs] You don't want to buy it, I'll buy it to myself." [laughs]

Mm-hmm

... by people.

Uh, maybe make it, uh, as mandatory as possible to, for people to register and, uh, maybe have approval process.

Second is possible to have the, some of endpoint, uh, which is also called the AI endpoint detection in the response.

Can we talk about some of the unique risks posed by the autonomous AI agents taking action across third-party networks, like identity permissions and tool execution? What are your big issues around the agentic AI security?

Yeah, I think the identity certainly is a big piece, right? Currently, there's no formal definition of what is agent identity, right? Definitely, if you ask 10 people to have 11 viewpoints of what identity is.

So is this human identity? Certainly not.

What cloud identity or machine identity is API or service account? It's also not, right? So that's why it's kind of really should be blended identity.

Blended identity with intent, I would say.

That's a good one.

A blended identity with intent.

It could be something else that you think is going to be kind of earth-shattering or interesting.