Permission Is Not Proof
The missing decision layer in enterprise agent platforms.
Watch: why agent credentials are not enough
Watch the episode with transcript →
Companion explainer recorded before the October 26 architecture response. The spending scenario is illustrative; this video is not a walkthrough of the full nine-page paper.
Permission establishes whether an action may proceed. Decision integrity asks whether it should proceed. A source-native receipt verifies what actually happened.
This nine-page paper responds to Naresh Waswani’s From AI Agent to Enterprise Agentic Platform. It develops a proposed decision-integrity layer alongside platform execution controls.
Download the full paper (PDF) →
Read the illustrated paper online →
What the paper covers
- A Governed Decision Object connecting claims, evidence, uncertainty, authority and action.
- Freshness, contradiction and human decisions at the execution boundary.
- Delegation limits and reapproval when the context changes.
- A compact technical appendix with an example object, execution rules and proposed acceptance checks.
Human decision integrity remains central.
The paper presents architecture and proposed acceptance criteria. Specific implementations must earn their proof.
By Jimmy Standaert · RiGi Group / Whitewater AIOS · CAIO / Forward Deployed Engineer/Architect
View the published LinkedIn paper and discussion →
Jimmy@RiGiGroup.com · 913-221-3137
rigigroup.com · whitewater.ai