Bradon Rogers

speaker

577 appearances 1 recordings 1 series first heard Jul 2026 last heard 13 Jul

Bradon Rogers’s voice in public audio — every appearance, attributed to the second.

Trend

recordings per month · last 12 months

Recordings per month over the last 12 months — 1 in all, peaking in Jul 2026 with 1.

Appearances

newest first · ▶ plays the moment

Eye On A.I. · Inside the Enterprise Browser Rebuilding Security for the AI Era | Bradon Rogers, Island · 13 Jul 2026 podcast

The agent is working over an application.

There's hidden instruction buried in the application.

The agent goes and follows those instructions, thinking those are the human giving them instructions in the process.

And then the next thing you know, the agent's over here grabbing tokens, corporate tokens out of a Gmail account that were sent via email.

And it shouldn't be doing that.

But at the end of the day, it's doing it and it's following what it thinks are instructions.

We've got users that are very basic level type stuff.

Users taking corporate data,

launching a personal app, AI app, and taking corporate data over to that personal AI app in a place we don't really want it because we just lost custody of our data because now it's gone into that AI provider's ecosystem.

It comes from a lot of different extension ecosystem.

You know, there's, I think I read the other day, there's like 18,000 plus classified extensions that are classified as AI.

That's a shadowy worm of things that wind up in your environment.

You don't even realize they're AI living in your environment.

There's another angle we've got to think about how we govern.

So there's all kinds of risks.

Again, a lot of it's centric around the data that you have in your organization.

But in the agentic world, it's also centered around what the agent may do autonomously on your behalf, thinking it's doing the right work.

and inadvertently doing something it shouldn't be doing.

So we want to build guardrails and policies around those agents, just as we have around the end users.

Again, a very smooth experience, but we want to let the user also, by the way, because the agent doesn't exist independent of a user.

Showing 141–160 of 577 · page 8 of 29