{"slug": "rubrik-ships-mcp-for-enterprise-ai-agents-co-engineered-with-anthropic", "title": "Rubrik Ships MCP for Enterprise AI Agents, Co-Engineered with Anthropic", "summary": "Rubrik has shipped a Model Context Protocol (MCP) implementation co-engineered with Anthropic, making it the third Fortune-500 vendor in two weeks to adopt MCP as a default agent interop standard after Salesforce and Google Cloud. Rubrik's version embeds data security into the protocol layer, aligning with the OWASP MCP Top 10 guardrails and replacing static API keys with Rubrik Agent Identity, which mints scoped, short-lived tokens per individual tool call. Co-Founder and CTO Arvind Nithrakashyap said the addition \"transforms security operations by seamlessly connecting customer AI agents directly into Rubrik's rich telemetry and governance framework.", "body_md": "Rubrik has become the third Fortune-500 vendor in two weeks to adopt the [Model Context Protocol](https://forkast.news/glossary/model-context-protocol-mcp/) (MCP) as a default agent interop standard. Following the lead of Salesforce and Google Cloud, as detailed in [Two Fortune-100 vendors just made MCP the default agent interop standard](https://forkast.news/two-fortune-100-vendors-just-made-mcp-the-default-agent-interop-standard/), Rubrik is signaling that the agent economy is moving beyond simple connectivity toward a more rigorous, security-first architecture.\n\nThe visible shift is adoption. The hidden shift is where the security burden lands. Rubrik enters this space by applying its existing data-security framework to the MCP layer, embedding identity and compliance directly into the interop protocol. This is qualitatively different from CRM or cloud vendors going agent-native – Rubrik is making the security layer itself agent-accessible.\n\nThe agent governance stack is formalizing into three distinct layers. At build time, [Cisco’s policy enforcement](https://forkast.news/cisco-ships-build-time-policy-enforcement-for-agent-frameworks-extending-the-governance-stack/) embeds constraints before an agent reaches production. At runtime, tools like [WSO2 and NVIDIA OpenShell](https://forkast.news/wso2-agent-manager-ga-lands-runtime-authority-as-a-distinct-infrastructure-layer/) provide control planes and sandboxing. Rubrik now occupies the third pillar: data-security. This layering suggests that agent governance is not a single product but a multi-layered stack.\n\nRubrik’s implementation is notable for its explicit alignment with the OWASP MCP Top 10 guardrails. By addressing risks like token mismanagement, privilege escalation, and context injection, Rubrik provides a standardized security template that other infrastructure vendors would be wise to follow. The company is moving away from static API keys and shared service accounts, replacing them with Rubrik Agent Identity – a system that uses scoped, short-lived tokens minted per individual tool call, significantly reducing the risk of credential leakage.\n\nThe platform also federates with enterprise identity providers like Okta and Microsoft Entra ID. When an agent performs an action, it is tied directly to human user permissions, enforcing Role-Based Access Control (RBAC) parity. The agent cannot exceed the permissions of the underlying user – a critical requirement for enterprise-grade security.\n\nArvind Nithrakashyap, Co-Founder and CTO of Rubrik, noted the strategic shift: “We are seeing rapid growth with AI, and the addition of Rubrik MCP transforms security operations by seamlessly connecting customer AI agents directly into Rubrik’s rich telemetry and governance framework.”\n\nOne practical benefit is the new agent inventory feature. By discovering and cataloging every active agent, MCP server, skill, and plugin, Rubrik aims to eliminate “shadow AI” – the proliferation of unmanaged, invisible agents within the enterprise. This visibility is a prerequisite for any serious security posture in an agent-driven environment.\n\nThe industry now faces several questions: will other infrastructure vendors adopt the OWASP-aligned guardrail model? How will the [agent identity](https://forkast.news/glossary/agent-identity/) and [runtime authority](https://forkast.news/glossary/runtime-authority/) layers converge? And how will the stateless MCP specification perform as enterprises scale to high-volume, agent-heavy environments? Rubrik’s move confirms that data security is no longer an afterthought in the agent economy – it is becoming the foundation.", "url": "https://wpnews.pro/news/rubrik-ships-mcp-for-enterprise-ai-agents-co-engineered-with-anthropic", "canonical_source": "https://forkast.news/rubrik-ships-mcp-for-enterprise-ai-agents-co-engineered-with-anthropic/", "published_at": "2026-09-17 11:02:58+00:00", "updated_at": "2026-09-17 11:23:27.532322+00:00", "lang": "en", "topics": ["agent-protocols", "ai-agents", "ai-safety", "ai-infrastructure", "ai-products"], "entities": ["Rubrik", "Anthropic", "Model Context Protocol", "Salesforce", "Google Cloud", "Arvind Nithrakashyap", "Okta", "Microsoft Entra ID"], "alternates": {"html": "https://wpnews.pro/news/rubrik-ships-mcp-for-enterprise-ai-agents-co-engineered-with-anthropic", "markdown": "https://wpnews.pro/news/rubrik-ships-mcp-for-enterprise-ai-agents-co-engineered-with-anthropic.md", "text": "https://wpnews.pro/news/rubrik-ships-mcp-for-enterprise-ai-agents-co-engineered-with-anthropic.txt", "jsonld": "https://wpnews.pro/news/rubrik-ships-mcp-for-enterprise-ai-agents-co-engineered-with-anthropic.jsonld"}}