Rubrik Agent Cloud has launched to provide runtime security and recovery capabilities for Anthropic’s Claude Code and Claude Cowork developer tools. The security and AI operations company announced the new integration on June 23, 2026, to help organizations deploy autonomous AI agents with greater observability and control. This solution addresses security gaps that emerge when autonomous agents write and deploy code without direct human oversight.
As organizations increasingly adopt autonomous AI agents to write and deploy code, traditional DevSecOps controls face new challenges. Security risks such as rogue commits, repository ransomware, prompt injection, and intellectual property exfiltration can occur at rapid speeds. Consequently, organizations require specialized tools to monitor agent actions and reverse unauthorized changes to their code repositories.
Securing AI Workforces with Rubrik Agent Cloud
The new platform introduces the Semantic AI Governance Engine, which secures and controls autonomous agents in real time using intent-driven governance. In addition, the platform features an Agent Inventory tool that provides visibility into risk profiles, access permissions, and policy violations across all active agents. These capabilities allow security teams to maintain oversight without slowing down development pipelines.
Reversing Unintended Actions and Codebase Damage
Through the Agent Rewind feature, Rubrik Agent Cloud allows organizations to reverse unintended actions taken by custom or pre-built agents. Meanwhile, the codebase resilience feature maintains immutable snapshots of GitHub and Azure DevOps repositories. These snapshots are stored outside the primary repository, protecting them from compromised credentials and allowing one-click recovery if an agent force-pushes incorrect code.
Configuration Tracking and Drift Monitoring
The integration also backs up, tracks, and restores the configurations governing how Claude agents behave, including system prompts, tool permissions, and settings files. The system continuously monitors for configuration drift and flags unauthorized changes before they can propagate. When drift is detected, the platform uses healthy backup snapshots to orchestrate a targeted recovery rather than performing a full rollback.
Executive Perspectives on Agentic Security
The company highlighted the rapid adoption of autonomous technologies and the corresponding need for governance. Security leaders require tools like Rubrik Agent Cloud to maintain control when autonomous agents are authorized to act within enterprise environments.
“Rubrik Agent Cloud gives organizations a resilience layer for Claude, which allows them to see what agents can access, govern what they do, rewind their actions, and recover both the code and the agent’s own configuration when something is destroyed or tampered with.”
Anneka Gupta, Chief Product Officer, Rubrik
As autonomous agents become more deeply integrated into software development, securing the agentic workflow will remain a priority for enterprise IT departments. The collaboration between Rubrik and Anthropic represents a step toward establishing standardized safety guardrails for autonomous code generation. Organizations can now deploy these tools with the assurance that unintended actions can be quickly identified and reversed.
