Geordie links Claude Enterprise to security controls

Geordie has integrated its security and governance platform with Anthropic's Claude Compliance API, extending its coverage across Claude Enterprise.

The integration gives corporate security and compliance teams access to activity inside Claude Enterprise, including chats, uploaded files, projects and platform logs. Geordie will combine that information with endpoint telemetry it already collects from tools such as Claude Desktop, Claude Cowork and Claude Code.

The development reflects a wider shift in how companies oversee artificial intelligence tools at work. Rather than simply tracking whether staff are using a tool, security teams increasingly want to monitor how AI agents operate over time, what context they build up and what actions they take inside persistent workspaces.

Traditional security tools have focused on endpoints and device activity, leaving a gap around project content and conversations created within Claude Enterprise. Anthropic's compliance interface provides programmatic access to that material, which Geordie is now bringing into its system.

Wider visibility

The integration is intended to give organisations a single view of Claude-related activity across the workplace, from code written by an agent on a developer workstation to files, chats and project content generated by business teams in Claude Enterprise.

Imported data will be normalised alongside existing telemetry and subjected to the same governance rules used elsewhere in the platform. Those checks include behavioural analysis, drift detection and risk controls designed to identify and respond to potentially problematic agent behaviour.

Beam, Geordie's context governance layer, can also apply inline guardrails and real-time controls across the Claude surfaces covered by the integration. Alerts and reporting can then feed into existing security workflows used by enterprise teams.

Security focus

The announcement comes as large companies face growing pressure to show that internal use of generative AI systems is governed with the same discipline applied to email, messaging and collaboration software. Tools such as Claude are moving from experimental use into operational roles, where they can retain context across tasks, interact with files and support business processes over extended periods.

That creates new oversight questions for chief information security officers and compliance leaders. Monitoring a login or a software installation may show that an employee accessed an AI tool, but it does not reveal what the model discussed, which documents were uploaded, what projects were created or how an agent's behaviour changed over time.

By connecting directly to Claude Enterprise through the compliance API, Geordie is positioning its platform around that deeper layer of visibility. The goal is to help companies govern not only tool access, but also the activity and context that build up inside AI systems as they become embedded in day-to-day operations.

The integration is now available to enterprise customers. Geordie describes its platform as focused on AI agent security and governance, with an emphasis on monitoring autonomous decisions, toolchains and workflows.

Hanah-Marie Darley, chief AI officer at Geordie, said the shift reflects the growing operational role of large language model systems inside companies.

"Claude is becoming operational infrastructure for enterprise teams: a system that takes actions, accumulates context, and executes across complex workflows. Organisations need the same programmatic assurance for Claude that they already have for Slack or Google Workspace. The Claude Compliance API makes that possible, and Geordie is the security and governance layer that makes it actionable," Darley said.