Anthropic expands Claude security features… inspection feature now officially available

Anthropic expanded security controls for its enterprise Claude service, moving its Inference hooks feature—which checks prompts and tool calls against security policies—into general availability while releasing a Security Center and centralized multi-organization access management in beta. The update, announced in a product release, lets enterprise security teams audit Claude's configuration and data handling and unify access policies across organizations, with each feature at a different availability stage. The Security Center compares settings across Claude Chat, Claude Cowork and Claude Code against Anthropic's recommended configurations. The move matters as enterprises demand stronger governance and auditability before scaling AI assistants into sensitive workflows. What to watch next is when the beta Security Center and central management features reach general availability and how broadly they extend across Claude products and enterprise tiers.