Anthropic Launches Enterprise-Managed Auth to Secure MCP Connectors
Brief
Anthropic has introduced enterprise-managed authorization for Model Context Protocol (MCP) connectors, allowing organizations to centrally provision, govern, and revoke Claude’s access to workplace applications through their identity provider.
The capability removes a recurring obstacle in enterprise AI deployments: individual users previously had to approve each connector after an administrator enabled it.
Under the new model, administrators authorize a connector once, map access to existing identity provider groups and roles, and users automatically receive the approved connections when they first sign in to Claude.
Anthropic Launches Enterprise-Managed Auth
Anthropic said the feature supports Datadog, Notion, and Slack, as well as Asana, Atlassian, Canva, Figma, Granola, Linear, and Supabase. Exa, Miro and Zoom support is expected next.