PROCEDURE · CoreAGK-PRC-07 v1.0Free preview

    AI Discovery and Shadow AI Procedure

    How to find the AI already in use before you write rules for it: staff survey, expense and SaaS data, network discovery, a no-blame amnesty, and recording what you find in the AI system register.

    Template version
    AGK-PRC-07 v1.0
    Format
    Word · 22 KB
    Used by
    AI Governance Lead, IT
    When
    Weeks 1–2
    Document type
    Procedure (Mandatory)
    Owner
    [AI Governance Lead]
    Approved by
    [AI Governance Committee]
    Review cycle
    Annual
    Parent document(s)
    AGK-POL-02, AGK-POL-03
    Adoption tier
    Lite: Required · Standard: Required · Regulated: Required
    AGK-PRC-07_AI_Discovery_and_Shadow_AI_Procedure_v1.0.docx

    1. Purpose

    To find AI already in use at [CLIENT NAME] — before governance starts and on an ongoing basis — decide what to do with each finding, and move staff from unapproved (T0) tools to approved tools with minimal disruption, so that the AI System Register (AGK-TPL-01) and Connector & Tool Registry (AGK-TPL-02) reflect reality.

    2. Scope

    • AI tools, SaaS products with AI features, browser extensions, desktop apps, connectors and add-ons used for [CLIENT NAME] business, whether paid by the company, expensed, or free.

    • AI features newly enabled by vendors inside existing approved software.

    • All staff, contractors and temporary workers with access to [CLIENT NAME] systems or data.

    • Out of scope: AI systems already registered and reviewed under AGK-PRC-05 (they are reconciled, not rediscovered).

    3. Triggers

    • Adoption of the AGK (full discovery, once).

    • Light discovery every [quarter].

    • An incident or near miss involving an unapproved AI tool (AGK-PRC-04).

    • Merger, acquisition or onboarding of a new business unit.

    • A vendor announces or enables a new AI feature in software already in use.

    4. Discovery sources

    Use as many sources as are available. Combining two or three gives a reasonable picture; no single source finds everything.

    Source What it finds Owner Lite?
    SaaS discovery / CASB or secure web gateway logs Sanctioned and unsanctioned AI web apps in use, users, volumes, uploads [IT Security] If available
    DNS / proxy logs for known AI domains Visits to AI services, by device or user, over the last [90 days] [IT Security] If available
    Browser extension inventory (device management) AI extensions and sidebar assistants installed on managed browsers [IT Operations] Yes, if devices are managed
    Expense and corporate card data AI subscriptions bought individually or by teams (search vendor names and “AI”, “GPT”, “assistant”) [Finance] Yes
    App marketplace / OAuth grants in the productivity suite Third-party AI apps granted access to mail, files or calendars; scopes granted [AI Platform Administrator] Yes
    AI features inside existing SaaS AI features enabled by default or by admins in CRM, HR, finance, helpdesk, design and meeting tools [Application owners] Yes
    Staff survey What people use, for what, and with what data — including personal devices and accounts not visible in logs [AI Governance Lead] Yes

    Staff AI-use survey (short form)

    Keep it anonymous by default, under [5 minutes], and send it with the amnesty message.

    1. Which AI tools have you used for work in the last [3 months]? (list; include free and personal accounts)

    2. Do you use a personal or a company account for each tool?

    3. What do you mainly use them for? (writing, summarising, research, coding, analysis, images, meetings, other)

    4. How often? (daily / weekly / monthly / rarely)

    5. What kind of information have you put into them? (public only / internal documents / customer or staff personal data / financial or confidential / source code)

    6. Have you installed any AI browser extensions, desktop apps or meeting note-takers?

    7. Have you connected an AI tool to your work email, files, calendar or other systems?

    8. Which AI features have you noticed switched on in software we already use?

    9. What tasks would you most like AI help with if an approved tool were available?

    10. Anything that worries you about AI at work? (free text)

    5. Classifying findings

    Each discovered tool or feature shall be classified into one of four outcomes by the AI Governance Lead, using AGK-STD-A02 AI Data Handling Standard and the tool tiers in AGK-000.

    The rest of this section and 9 sections are in the download

    You have read about 31% of this document. The editable Word file comes with the free Lite pack.

    Get the full document

    Free AI Governance Lite pack

    The documents an organisation starting out needs to govern AI tools in its first 90 days. Word and Excel, ready to edit.

    • Lite Quick Start and Control Checklist AGK-001 · Word
    • AI Governance Framework and Document Architecture AGK-000 · Word
    • AI Acceptable Use Policy AGK-POL-02 · Word
    • AI Data Handling Standard AGK-STD-A02 · Word
    • AI Discovery and Shadow AI Procedure AGK-PRC-07 · Word
    • AI Adoption Roadmap and Use-Case Prioritisation Guide AGK-GDL-C02 · Word
    • AI System Register AGK-TPL-01 · ExcelPreview AI System Register
    • AI Use Case Intake Form AGK-TPL-03 · Word
    • AI Incident, Nonconformity and CAPA Log AGK-TPL-06 · ExcelPreview AI Incident, Nonconformity and CAPA Log
    • AI Acceptable Use Acknowledgement AGK-TPL-07 · Word
    • Pengakuan Penggunaan AI (Bahasa Melayu) AGK-TPL-07-BM · Word
    • AI Use-Case Prioritisation and Benefits Tracker AGK-TPL-12 · ExcelPreview AI Use-Case Prioritisation and Benefits Tracker

    Template release v1.0 · ZIP · 733 KB

    Download AI Discovery and Shadow AI Procedure and the Lite pack

    Your details for this download

    We use your details to send this download and to understand which documents are useful. Marketing emails only if you tick the box.

    Templates for governance purposes, not legal advice.

    Powered by EzyChat