Skip to content
Request a demonstration

Demonstration · AI governance, security and trust · 4 steps

From what the AI may read to answers with evidence

Set what an assistant may read, audit who has access, answer a customer's security questions with evidence, and check what the Trust area can honestly say.

Example: Brightwater Software (fictional) · Head of security and trust. Fictional data and a scripted answer.

What happened
A customer wants to use the company's AI assistant on its own data and has sent a security questionnaire.
The question
What may the assistant read, who can reach it, what can we honestly answer, and what do we say we do not hold?

The journey, step by step

Simulation. A scripted journey on fictional data, not produced by a live model. Each step uses what you decided in the steps before it, and only the figures are computed.

  1. What the assistant may readChoose the data sources the assistant may read. Restricted material stays out; an uncertain classification needs a person.Open

    Choose the data sources the assistant may read. Restricted material stays out; an uncertain classification needs a person.

    Choose what the platform may read. Restricted material cannot be put in scope here, and a classification the analysis is unsure of must be verified by a person.

    Read it?SourceHoldsSensitivityWhat the analysis thinks it is
    Published product documentation
    Public site    		540 documentsopenManuals and datasheets clear
    Internal procedures
    Quality system    		1,900 documentsinternalApproved procedures clear
    Support knowledge base
    Helpdesk    		3,200 documentsinternalSupport articles check
    Board minutes
    Governance system    		260 documentsrestrictedBoard papers clear
    Customer contracts
    Contract system    		840 documentsrestrictedCommercial terms clear

    Sources in scope
    0 of 5
    documents in scope
    0
    Restricted, kept out
    2 sources
    To verify
    0

    0 of 5 sources are open to the assistant, 0 documents.

    The assistant reads open and internal material and never restricted material [S1]. 0 included source(s) still need a person to verify their classification.

    Relies on: S1 · Data source inventory

    The head of AI governance decides. Approving sets what the assistant may read. It does not switch the assistant on, and it can be narrowed or revoked at any time.

    Not ready. Choose at least one source to read first.

Next: Who can reach it. It opens when the step above is complete.

Limits of this example

  • Nothing here claims a certification, an audit result or any assurance the company does not hold, and the journey shows absence as absence.
  • A scripted journey on fictional data.
  • This is a scripted demonstration on fictional data. No model was called and no real organisation, person or result is shown.

The AI governance, security and trust story · All demonstrations

The other journeys