Docs
relyance.ai ↗
Browse docs

FAQ: AI Use Cases

Updated November 5, 2026View as Markdown ↗

Audience: prospects evaluating Relyance, and new customers getting started.

What is an AI use case, in one paragraph?

An AI use case is a documented AI system in your organization - a business purpose with an accountable owner and a risk level - built from the models, AI agents, MCP servers, and vendors Relyance has discovered in your environment. It's the level at which regulators and frameworks such as the EU AI Act, the NIST AI RMF, and ISO/IEC 42001 reason about AI, so it's where assessments, risks, policies, approvals, and evidence come together. See the Overview and Implementation Guide.

Where do I find it?

Under Observe → AI, on the Use Cases tab.

Is it available to every tenant?

Not yet. AI Use Cases is in Beta and is available to select customers. Change management - approval, change detection, and lifecycle stages - is enabled separately. Contact your account team for access.

Do I have to fill in every use case by hand?

No. Describe the use case in a sentence or two, optionally with a design doc attached, and Relyance drafts it - name, description, department, owner, risk level, custom fields, and matching components from your inventory - for you to review. You can also start a use case from a model, AI agent, or MCP server you've found in the inventory. See Creating a use case.

What's the difference between a use case and a model or agent?

Models, AI agents, and MCP servers are components that Relyance discovers. A use case is what your organization does with them. One component can be part of several use cases, and a use case can use several components. A change to a shared component shows up on every use case that uses it.

Why does my use case say Unregistered?

A use case is Registered only when every attached model, AI agent, and MCP server is registered. One unregistered component makes the whole use case Unregistered. Hover Governance Status in the use case drawer to see which components are holding it back, then register them on their own tabs. A use case with no components, or only vendors, is always Unregistered.

Do vendors count toward governance status?

No. Vendors are listed among a use case's components, and their policies and assessments still apply, but they don't count toward governance status or usage status.

Which assessments run on a use case?

When you create a use case, Relyance can start the AI Use Case Intake assessment (a short triage) and the full AI Governance and Risk Assessment. Both are ticked by default. You can also run any other template against a use case. See Starting assessments when you create a use case.

How do I avoid cataloging the same AI system twice?

Relyance compares each new use case with your existing ones. If it looks like one you already have, you can compare them and add to the existing use case instead. Each use case also keeps a Similar Use Cases card. See Similar use cases.

How do risks get onto a use case?

Two ways. Risks raised from the use case's completed assessments are linked to it in the Risk Register automatically, and you can apply any register risk to a use case. Policy violations against its components appear as issues. Both show on the use case's Risks & Issues card.

What happens when an approved AI system changes?

Where change management is enabled, Relyance compares every scan with what you approved, scores each change, and accepts low-risk ones on its own with an explanation. Higher-risk changes wait for a person, and accepting one reopens the intake questions it affects. See Change Management and Lifecycle.

Can I approve a change before it ships?

Yes. Submit a change request from the use case's Changes tab. You see the risk it would raise before you submit, and once it's accepted, the change counts as approved when a scan sees it ship. See Change requests.

Does quarantining a use case stop the AI system?

No. Quarantine flags the use case in Relyance to be stopped and investigated, but it doesn't switch anything off in your environment.

What do I give an auditor?

Export a compliance package from the use case drawer: one ZIP with the use case report, a report for each attached model, assessment reports, the use case's risks and issues, and its full audit log. See Exporting a compliance package.

What happens to assessments if I delete a use case?

They're kept, but any assessment that named the use case as its subject is left without one. The delete confirmation tells you how many are affected. Deleting a use case can't be undone.

Who can create and edit use cases?

Anyone who can see Observe → AI can view use cases. Creating, editing, deleting, approving, and accepting changes need edit access to the platform.