Interview of Tim on SiliconANGLE theCUBE

AI agents now access systems and take action on their own, so the next incident may not start with an attacker. An agent with legitimate access can still act on bad context and cause damage at a speed traditional controls weren't built for.

Timothy Youngblood, Chief Strategy Officer at Onyx Security and a four-time Fortune 500 security chief, explains how to get ahead of it:

• Why runtime control matters more than logs when agents act in seconds.

• How to find excessive privileges hidden in agent identities.

• What a graded kill switch looks like, so you contain risk without stopping the business.

• How to reconstruct what an agent did, and who owns the response.

• Which tabletop exercise to run now.

Watch the full conversation to learn how to treat AI agents as your newest insider risk.