News
Rubrik Debuts 'Agent Cloud' to Monitor and Govern AI Agent Actions
Rubrik today introduced a new enterprise platform called Rubrik Agent Cloud, aimed at helping organizations deploy AI agents while maintaining operational control. The company frames it as the first enterprise control layer for the AI agent lifecycle, extending its cyber recovery heritage into agent operations.
In announcing the new offering, CEO Bipul Sinha described the agentic AI revolution now taking place, where AI agents can operate autonomously: analyzing data, making decisions, coding, resolving support issues, and offering recommendations. But they also introduce new risks, he indicated, from unintended actions to potential compromise, with serious consequences for operations and reputation.
"At Rubrik, we believe progress should never come at the cost of control," he said. "That belief is what led to Rubrik Agent Cloud."
[Click on image for larger view.] Risks and Violations (source: Rubrik).
It builds on previous Rubrik tech found in the company's Security Cloud, now bringing together data, identity, and application context to help organizational leaders answer questions such as:
- What agents do I have?
- What are they capable of doing?
- What data can they access?
- What did they just do?
- And if something goes wrong, can I undo it?
Regarding that latter question, the company previously addressed it with
Agent Rewind, announced in August (see the
Virtualization & Cloud Review article, "
Rubrik Hits 'Undo' on AI Agent Mistakes").
According to Rubrik, the new platform does these three things:
First, discovers and maps agents from popular builders and platforms and maintains immutable audit trails with data, identity, and application context. The announcement cites auto-discovery across OpenAI, Microsoft Copilot Studio, Amazon Bedrock and others, with continuous monitoring of activity and data access. The product page summarizes this "Agent Monitor" capability as: "Find agents automatically. See them in one place. Keep tabs on how they interact with your data, identities, and applications, so you can spot issues early and resolve them with confidence."
Second, applies policy enforcement to keep agents within approved boundaries. Rubrik's "Agent Govern" is described as defining and enforcing agent behavior, access, and action policies in real time, with integration to enterprise identity systems for secure and compliant operation. The product page summarizes: "Define how agents behave, what they can access, and when. Apply policy-based controls in real time so agents operate safely and within approved boundaries."
Third, enables rollback of unwanted or destructive actions, as noted above. The platform integrates with Rubrik Security Cloud to provide selective time and blast-radius rollback "without any downtime or data loss." The product page mirrors that positioning for "Agent Rewind": "Recover fast when AI agents make mistakes. Agent Rewind helps you undo unwanted or destructive actions without downtime or data loss."
Rubrik connects Agent Cloud to broader enterprise readiness themes. The product page highlights concerns it says are common in the field: 90% of IT leaders worry about shadow AI, 63% lack confidence in recovery when AI goes wrong, and 73% struggle with data governance and compliance.
The cyber resilience and data security specialist is offering Agent Cloud via a waitlist on the product page. The announcement post includes a Safe Harbor notice and states that "not all features are currently available."
The company will reveal more details in a Nov. 6 webinar.
About the Author
David Ramel is an editor and writer at Converge 360.