Know What AI Is Running Across Every Managed Endpoint

AI tools are being installed across managed environments faster than IT teams can inventory them. N-central and N-sight now detect and classify AI tools as a distinct category, directly in Assets View, with no new agent and no new deployment.

• Employees install ChatGPT desktop apps, Copilot and Claude IDE extensions, browser plug-ins, and local models — often without approval or governance.
• IT has no visibility into what’s running.
• Proprietary data, customer records, and source code can flow into tools entirely outside your control.

What’s new

A continuously updated AI inventory.

The existing endpoint agent now identifies every AI tool installed across managed devices and classifies each one by category (LLM, code assistant, image generation, transcription, and more), vendor, and model family. AI is treated as its own detection category.

• LLMs, code assistants, image generators, transcription tools, and more are identified and classified automatically.
• Install state is tracked over time, not as a point-in-time snapshot. You can see when tools appeared and how usage is shifting across your environment.

Visible where you already work.

• A new AI Services column in Assets View shows which devices are running which AI tools at a glance.
• A dedicated AI Services tab in Asset Details lets you drill into specifics: service name, vendor, detection method, process state, and file path.

Governed data across every surface.

• AI signals are exposed through the GraphQL API, MCP Server, and N-zo, so your existing queries, automations, and AI-driven workflows can act on them immediately.
• Same data, every surface, no manual export.

 

Why this matters now

Auditors are starting to ask for AI inventories. Cyber insurers are adding AI governance questions to renewal applications. Compliance frameworks increasingly expect evidence of AI oversight.

This gives you the answer: a structured, audit-ready record of what AI is running, where, and on which devices.

• Enforce acceptable-use policies with data, not assumptions.
• Meet data-residency and compliance obligations with evidence you can produce on demand.
• Prove exactly what AI is running when auditors, insurers, or leadership ask.

What’s ahead

This release covers AI tools installed on endpoints: desktop apps, IDE plugins, browser extensions, CLIs, and portable executables.

A significant portion of AI usage also happens over the network through browser-based services and API-level access that endpoint-only discovery does not reach. Future releases will expand detection in this direction.