Brief / AIGIS resources
Most AI tools log prompts and responses. That answers what was said, not whether the user was allowed to see the underlying data. AIGIS records the permission decision before model exposure. Salesforce is the production proof path.
Executive read
Prompt logs capture output, not access rights.
An auditor asks why the AI was allowed to know a record, not just what it returned.
Permission-provenance records the checks and the stripped fields per interaction.
Prove it on Salesforce first; disclose ServiceNow and SAP as co-development.
Analysis
A transcript shows the answer. It does not show whether the requesting user had object, field, and record access to the data behind that answer.
If two users ask the same question and should see different data, the transcript alone cannot prove the boundary held.
AIGIS checks permissions before the model sees business context, strips fields the user cannot view, and records that decision as auditor-readable evidence.
That record is what lets a reviewer confirm the access boundary was enforced, not assumed.
See a real Salesforce workflow with the provenance record at `/demo`.
Resource packet
Evidence packet
Capture user context, system of record, enforcement tier, stripped fields, model route, response, hash marker, and fallback notes.
Salesforce is the production proof path. ServiceNow and SAP are design-partner co-development paths with asymmetric enforcement that must be disclosed in diligence.
Keep reading
Audit
The audit elements needed to explain what an AI answer was allowed to know.
Audit
What audit teams should expect from AI permission-provenance evidence across Salesforce-first governed workflows.
Architecture
Why enterprise AI governance should check permissions and remove inaccessible context before any model call.