Deploy AI Agents Faster & Safer with Rubrik
▼ Summary
– Rubrik launched Agent Cloud to help enterprises adopt AI agents while managing deployment risks like hallucinations and security threats.
– The platform monitors agent activity, enforces real-time behavior controls, and maintains immutable audit trails for accountability.
– It provides tools to undo destructive agent actions through selective rollback capabilities without downtime or data loss.
– Agent Cloud integrates with enterprise identity systems and popular AI builders like OpenAI and Microsoft Copilot Studio for secure operations.
– The solution is currently available through limited early access, with some features still in development for future release.
Businesses today face a dual reality where AI agents present both immense potential and significant operational hazards. Rubrik, Inc. has introduced the Rubrik Agent Cloud, a platform engineered to speed up the adoption of enterprise AI agents while effectively managing the risks tied to their deployment. This solution arrives at a time when AI transformation is essential, yet IT leaders grapple with serious concerns like AI hallucination and the threat of malicious compromise.
The Rubrik Agent Cloud is purpose-built to monitor and audit agent activities, enforce real-time behavioral guardrails, fine-tune agents for improved accuracy, and—critically—undo agent mistakes. It leverages the Rubrik Platform, which integrates data, identity, and application contexts, to deliver security, precision, and operational efficiency as companies evolve into AI-driven enterprises.
Bipul Sinha, CEO, Chairman, and Co-Founder of Rubrik, emphasized the urgency, stating, “IT and security teams frequently lack visibility into what their AI agents are doing or how to reverse errors. We help answer fundamental questions: What agents are deployed? What can they do? How are they performing? What actions have they taken? And can we undo problematic changes? The impact of an AI agent mistake can be ten times worse in a tenth of the time. Our Agent Cloud directly confronts this by applying our expertise in data, identity, and resilience so customers can deploy AI with confidence.”
The platform delivers comprehensive agent management spanning the full lifecycle—from observability and governance to performance tuning and simulation.
Agent Monitor automatically discovers infrastructure-as-a-service agents (such as those on Azure or AWS) and platform-as-a-service agents (like M365 or AgentForce). It identifies and maps active agents across leading development tools including OpenAI, Microsoft Copilot Studio, and Amazon Bedrock. The system continuously watches agent activity and data access, maintaining immutable audit trails that capture context from data, identity, and applications.
Agent Govern tracks how agents are used and evaluates their performance against given prompts. It provides teams with tools to control destructive or unwanted actions, defines and enforces real-time policies for agent behavior and access, and offers a centralized interface for integration with enterprise identity systems—ensuring innovation remains secure and compliant.
Agent Remediate, announced for August 2025, integrates with Rubrik Security Cloud to deliver precise rollback capabilities for undesirable or harmful agent actions. This feature moves beyond simple monitoring by allowing organizations to instantly reverse unwanted changes without causing downtime or losing data. Selective rollback of agent-driven modifications guarantees continuous protection for vital data and systems through immutable recovery processes.
Rubrik Agent Cloud is currently accessible via a limited early access program for selected customers, with certain features still in development. To learn more about advancements in agent operations, visit the Rubrik blog or register for an upcoming webinar.
(Source: ITWire Australia)
