Demo Corporation/AI trust center
Summary (PDF)

AI trust center

How Demo Corporation uses AI

Demo Corporation uses AI tools to write, summarise, code and answer customer questions faster. We know customers want to understand what that means for their data, so this page states it plainly.

Our people may use only the AI tools listed here, each under a business agreement. We monitor which AI tools are used on company devices, never what is typed into them. The facts below are read from our governance records each time this page is published.

Demo Corporation is a fictional company. This page is an example of a GhostAudit AI trust center.

AI policy
In placeApproved 7 Oct 2026
Approved AI tools
7Listed by name below
Customer data used to train AI
NoConfirmed 3 Oct 2026
AI usage monitoring
Metadata onlyNo prompts or content

How we govern AI

As of 10 Nov 2026
AI acceptable use policy
In placeApproved 7 Oct 2026
Policy review
Once a yearNext review due 10 Nov 2027
Accountable for AI governance
Head of IT and Security
AI inventory
MaintainedA register of the AI systems in use, with an owner and a risk level for each

Approved AI tools

As of 10 Nov 2026

These are the AI tools Demo Corporation has approved for company work. Tools that have not been approved are not permitted.

  • Adobe FireflyImage generation
    Adobe
  • ChatGPTAI assistants
    OpenAI

    Drafting, summarising and research across all teams. Enterprise workspace with single sign-on.

  • ClaudeAI assistants
    Anthropic

    Long documents, analysis and customer support drafts. Team workspace with single sign-on.

  • GitHub CopilotCoding assistants
    GitHub (Microsoft)

    Code completion and review for the engineering team.

  • GrammarlyWriting assistants
    Grammarly

    Spelling, grammar and tone in written communication.

  • JasperContent generation
    Jasper AI
  • Microsoft CopilotAI assistants
    Microsoft

    Email, meetings and documents in Microsoft 365, signed in with the work account.

Customer data and AI

  • Is customer data entered into AI tools?

    Only into the approved AI tools named below, for defined business purposes.

    Support and success teams may use customer account context to draft replies. Bulk exports of customer data are not permitted in any AI tool.

    Confirmed by the workspace owner on 3 Oct 2026

  • Which AI tools may process customer data?

    Only these approved tools:

    • ChatGPT
    • Claude
    • Microsoft Copilot

    Confirmed by the workspace owner on 3 Oct 2026

  • Under what contractual terms?

    Business or enterprise agreements that include a data processing agreement.

    Confirmed by the workspace owner on 3 Oct 2026

  • Is customer data used to train AI models?

    No. Our agreements with these vendors exclude training on our data.

    This is also switched off in each workspace's admin settings.

    Confirmed by the workspace owner on 3 Oct 2026

  • Is AI output reviewed by a person before it reaches customers?

    Yes. A person reviews AI output before it is used in anything sent to a customer.

    Confirmed by the workspace owner on 3 Oct 2026

Monitoring and controls

As of 10 Nov 2026
  • Use of AI tools on company devices and browsers is monitored.
  • Monitoring records metadata only: which tool, when and for how long. Prompts, page content, keystrokes and screenshots are never collected.
  • A tool policy defines which AI tools are approved, under review or prohibited.
  • Employees are warned before they use an AI tool that has not been approved.
  • Prohibited AI tools are blocked on managed devices.
  • On-device checks warn before sensitive data is submitted to an AI tool. The check runs on the device; the content never leaves it.
  • Usage records are removed after 1 year.
  • A process is in place to record, contain and review AI incidents and near misses.

Frameworks and assessments

As of 10 Nov 2026
FrameworkLast assessed
EU AI Act31 Oct 2026
ISO/IEC 4200121 Oct 2026
NIST AI Risk Management Framework11 Oct 2026
SOC 2 (AI-relevant criteria)11 Sep 2026

These are self-assessments of AI systems against each framework, carried out in GhostAudit. They are not certifications or independent audits.

Employee training and policy acknowledgement

As of 10 Nov 2026
  • Employees are asked to read and acknowledge the AI acceptable use policy.

Documents available on request

As of 10 Nov 2026

Demo Corporation shares these documents with customers and prospects on request.

  • AI acceptable use policyOn request

Request access

Demo Corporation reviews each request. If it is approved you receive a time-limited link to the documents, marked with your name.

Your details go to Demo Corporation to decide on this request and are not used for anything else.

Contact

Questions about how we use AI, or a security questionnaire to complete? Write to our security team.

[email protected]

Change log

    • Added Claude to approved AI tools
    • Updated the answer: Which AI tools may process customer data?
    • Added section: Frameworks and assessments
    • Added section: Documents available on request
    • Trust center published