BREAKING from Rubrik!!!! They just made the most aggressive bet I have seen on where enterprise security is heading. I interviewed Anneka Gupta, their Chief Product Officer, right as it all went public at Rubrik FORWARD on The Ravit Show.

Two announcements came out of Las Vegas this week.

First, Rubrik AI. The platform itself is now an agent. You define the outcome, recover clean, contain the blast radius, restore the business, and Rubrik AI reasons over your data, identities, and deployed agents to deliver it. Recovery sequences that took human teams weeks now finish in minutes. Every action stays auditable, attributable, and reversible.

Second, Rubrik Agent Cloud for Anthropic's Claude Code and Claude Cowork. Claude is being adopted faster than any agentic technology Rubrik has seen. These agents write, push, and deploy code on their own, while enterprise security was built assuming a human stays in the loop. RAC closes that gap with real-time governance through SAGE and the industry's only Agent Rewind, which reverses an agent's actions and recovers the codebase even when a mistake outruns version control.

Here is why these two launches are really one story.

Rubrik's Zero Labs research found 86 percent of firms expect AI agents to outpace their existing security capabilities. Most vendors respond to that stat by selling more visibility. Rubrik's answer is different: if threats and agents move at machine speed, defense and recovery have to move at machine speed too. So they built an agent to protect you from agents.

That framing is what I pushed Anneka on in our conversation.

We got into what a runaway AI risk actually looks like inside a security environment, and how Agentic Guardrails stop one before it spreads. Which parts of a multi-week recovery workflow are genuinely automated and which still need a human call. How one agent reasons across Rubrik Security Cloud and Rubrik Agent Cloud at the same time, spanning data, identity, and third-party agents. The role of identity in agentic resilience. And why Databricks Unity Catalog was chosen as the first native lakehouse integration for RAC, with more connectors coming.

My take after 750 plus interviews in this space: every enterprise I talk to is racing to deploy agents, and almost none of them can answer one question. What happens when an agent does something wrong?

Observability tells you what happened. Rewind lets you undo it. That difference is going to define the next phase of enterprise AI, because the companies that win with agents will not be the ones that deployed fastest. They will be the ones that stayed in control.

#data #ai #cybersecurity #theravitshow