Shadow SaaS and AI Vendor Discovery Checklist

An editable 17-page field guide for the analyst who has just found an unapproved tool in the finance data, typically an AI note-taker someone expensed. Discovery is the easy half. The work is establishing who uses it, what data has already been entered, what it integrates with, whether the content trains a model, and then choosing between approving, restricting, migrating or removing it. Every item explains why it matters, the result a good answer should produce, the evidence to request, and the next analyst move.

What’s inside

• A readiness page confirming the vendor entity, product, environment, region and service in scope, business criticality and data sensitivity, the internal owners who can validate each claim, and how restricted evidence will be handled
• 30 numbered items, SSA-01 to SSA-30, grouped into 5 control domains and classified as they are in the guide: 23 marked Core, 2 marked Elevated and 5 marked AI Specific
• Governance and Scope; Discovery Signals and Coverage; and Ownership and Exposure Validation
• AI Specific Risk; and Response and Continuous Control
• Response, evidence reference, action owner and due date against every item, so the completed form is the assessment record rather than a questionnaire you transcribe later
• A rapid final checklist that restates all 30 items on one page for the closing challenge

The assessment rule is written into the form: evidence should be current, complete for the service in scope, usable by your organization, tested where capability matters, and linked to a decision. A certificate, policy, screenshot or vendor statement is not enough on its own when it does not prove the specific requirement.