Every model on one register
Purplecop AIMS records each model with its owner, its purpose and its risk classification.
Model inventory
Purplecop AIMS · Model inventory
Every AI model and agent, accounted for
AI arrives team by team: a model here, an agent there, a tool bought on a card. Without one register, nobody can say what AI the company runs or how risky it is. Purplecop AIMS keeps that register, with an owner and a risk score for every entry.
The difference
From scattered AI to a governed register
How an AI agent becomes something you can account for.
Why Purplecop
What sets the AIMS register apart
AI governance on the same library as your security frameworks.
AI risk scored on autonomy
Model risk is scored on impact and autonomy, not likelihood alone, so an agent that acts by itself is weighed differently.
AI risk register
Vendor AI on the same scale
Third-party models and AI services are assessed like any other vendor.
AI vendor assessment
ISO 42001 and EU AI Act in the library
140 ISO 42001 and 279 EU AI Act requirements are seeded and linked to controls.
Frameworks
AI policies from the AI Policy Builder
Acceptable use, model lifecycle and oversight policies to start from, tagged to the AI frameworks they serve.
AI Policy Builder
Internal AI audits
Audit the AI management system internally, with the evidence attached, before an external auditor does.
AI audit
What you get
Built on one platform, with people who use it
Platform
Purplecop AIMS
Model inventory, AI risk, AI controls and the evidence that an AI system is governed, mapped to ISO 42001, the NIST AI RMF and the EU AI Act.
Model inventoryAI risk registerAI vendor assessmentAI frameworksAI policies
Services
Practitioners beside the platform
The team that runs your engagement works inside the same system, so the work they do stays with you.
Also on the platform
TPRM
Onboard, assess, score and continuously monitor every vendor in one workflow.
Third-party risk
Asset Intelligence
Asset discovery, network monitoring and the audit log in one module, so every asset’s data lives in one place.
Asset Intelligence
Phishing Simulation
Live phishing campaigns with per-user risk scores, run entirely in-house.
Human layer
LMS
Role-based security awareness training for every employee, tracked to completion.
Human layer
People
A directory of the people in scope for training, policy sign-off and access reviews.
People
RBAC
Role-based access control with granular permissions, on a least-privilege model.
Access control
Questions
What teams ask us about AI agent inventory
Does AIMS find our AI agents automatically?
No. Models and agents are registered by their owners; the register then keeps them scored and mapped.
How is AI risk scored?
On impact and autonomy, not likelihood alone.
Does it cover vendor AI?
Yes. Third-party models and AI services are assessed on the same scale as any vendor.
How much faster is it?
It scales with how much you integrate: the more connectors you turn on, the more evidence is collected automatically, saving up to 70% of the time you used to spend preparing for each AI review.
See it against your own scope
We will walk the product against your scope and show you what runs on day one and what needs configuring.