At its Rubrik Forward conference, Rubrik unveiled significant updates transforming its data security platform into an AI agent-first system. The company also made generally available its Agent Cloud tailored for Anthropic’s Claude, tackling rising risks from autonomous AI agent actions in enterprise environments.
- Rubrik transforms its platform into an autonomous AI agent with agent-first interface.
- Agent Cloud for Anthropic’s Claude now generally available with real-time governance.
- New partner program to deploy Agent Cloud for Claude via global systems integrators.
What happened
Rubrik Inc. announced at its Rubrik Forward conference a major upgrade turning its data security platform into an autonomous AI agent. This agentic platform features a patent-pending design that spans across Rubrik Security Cloud and Agent Cloud, enabling multi-step recovery processes that were previously slow, while ensuring all autonomous actions remain auditable, attributable, and reversible. Alongside this, Rubrik made its control layer for Anthropic’s Claude AI publicly available as part of its Agent Cloud service.
The Claude-enabled Agent Cloud introduces a Semantic AI Governance Engine (SAGE) that applies intent-driven controls to autonomous agents in real time. It includes unique tools such as Agent Rewind to reverse unintended agent actions, Codebase Resilience for continuous immutable backups of code repositories, and a configuration version tracker to detect and prevent malicious changes to an agent’s behavior. Rubrik also introduced Project Hourglass, a partner initiative with six global systems integrators to promote and embed the Claude Agent Cloud within enterprise AI security frameworks.
Why it matters
As enterprises increasingly deploy AI agents to automate critical workflows including code development and deployment, conventional security models assuming constant human oversight are insufficient. Malicious actors or compromised agents can execute destructive actions at machine speed, risking data integrity, intellectual property, and operational continuity. Rubrik's AI-first approach addresses this vulnerability by embedding agentic automation that monitors, governs, and can rapidly recover from harmful autonomous activity.
The availability of the Agent Cloud for Claude offers enterprises specifically targeted protections for AI-driven code operations, where ransomware, unauthorized commits, and prompt injection pose growing threats. By continuously backing up repositories and configurations outside primary systems, Rubrik ensures that organizations can restore trusted code and agent settings quickly, minimizing downtime and potential damages. This enhances confidence in AI adoption by securing critical DevOps environments from dynamic and fast-spreading AI threats.
What to watch next
Adoption of Rubrik's new AI-driven platform and Agent Cloud for Claude will be a key indicator of enterprise readiness to govern increasingly autonomous AI agents. The success of Project Hourglass and its partners like Cognizant, Deloitte, and Wipro in integrating these offerings into their AI security frameworks will influence broader market acceptance.
Future developments should be monitored around the evolution of Rubrik AI’s agentic capabilities and how effectively it mitigates both internal misconfigurations and sophisticated external AI attacks. Enterprises may also watch for further integrations with other AI systems beyond Claude, as well as enhancements to audit and recovery processes to keep pace with emerging AI threat vectors.