Rubrik’s New Cloud Service Enhances Security for AI Agents

In an era where AI agents are becoming increasingly autonomous, security concerns are paramount. Rubrik, a key player in the security and AI operations sector, has introduced its Rubrik Agent Cloud (RAC), designed specifically for Anthropic’s Claude Code and Claude Cowork. This new service provides organizations with the tools to deploy AI agents at scale while ensuring effective oversight and control.

The launch of RAC is timely, given the rising threat of rogue commits and other security risks. Traditional enterprise security frameworks often rely on human oversight, leaving gaps when AI operates independently. With AI agents now capable of writing, pushing, and deploying code autonomously, the potential for cyber threats has grown significantly. Rubrik’s solution seeks to address these risks with features like Agent Rewind, which allows organizations to reverse unintended actions and recover from incidents that may occur swiftly, often beyond the reach of standard DevSecOps controls.

RAC includes several key features aimed at enhancing the security and resilience of AI agent operations. At the forefront is the Semantic AI Governance Engine (SAGE), claimed to be the first of its kind. This tool offers real-time oversight of autonomous agents, moving from static, manual processes to dynamic, intent-driven governance. This shift is essential for organizations looking to expand their AI workforce without sacrificing security.

Key Features of Rubrik Agent Cloud

Comprehensive Visibility and Control

RAC features an Agent Inventory tool that provides organizations with extensive visibility into potential risks, access permissions, and policy violations across all deployed agents. This comprehensive view is crucial for managing the complex interactions AI agents can have within an enterprise.

Intelligent Recovery Mechanisms

Another notable feature is the enhanced Codebase Resilience, which maintains immutable snapshots of GitHub and Azure DevOps repositories. This capability allows organizations to recover from actions that cannot be undone through traditional version control methods. For example, if an AI agent were to force-push over commit history or delete branches, RAC enables one-click recovery to a known-good state, including options for ransomware rollback. Such features highlight the need for a recovery strategy that goes beyond human intervention.

Monitoring for Configuration Drift

Rubrik’s service also continuously monitors for configuration changes across agents. This proactive approach allows the platform to detect and flag malicious or unauthorized changes before they impact operations. By linking detected drifts to healthy backup snapshots, Rubrik ensures that organizations can quickly recover from potential threats without significant downtime.

Expert Insight

Anneka Gupta, Chief Product Officer at Rubrik, stressed the importance of such solutions in today’s digital landscape. "As AI agents evolve and take on more responsibilities, their security must be a top priority for enterprises. Our Rubrik Agent Cloud is designed to provide the necessary layers of control and resilience for organizations deploying Anthropic’s Claude," Gupta stated.

Looking Ahead

The rollout of Rubrik Agent Cloud marks a significant development for enterprises utilizing AI agents. By incorporating advanced security measures into the operational framework of AI, Rubrik addresses critical vulnerabilities that have surfaced alongside the rise of autonomous technologies. As organizations continue to navigate the complexities of AI integration, solutions like RAC will be vital in protecting their digital assets and ensuring compliance in a rapidly evolving environment.

The demand for stable security measures in AI operations has never been more urgent. With Rubrik’s new cloud service, organizations can deploy AI agents with greater confidence, armed with the tools to effectively manage risks and maintain operational integrity.