Claim, evidence,
and context.
The book examines how generated output becomes warranted knowledge.
Explore the vocabularyA book about governing AI in the enterprise
Ottaviano De Cicco
Intelligence,
with a mandate.
A framework for governing what your organization knows, decides, and allows AI agents to do.
Inside the framework Four dimensions · three planes · three loops+
Four separate decisions
- Standing
- Is this claim warranted for this use?
- Persistence
- What remains eligible for use, and what history must be kept?
- Visibility
- Who may retrieve which representation and context?
- Action enablement
- Who may act, within which boundary?
Three control planes
Knowledge and Authority supply distinct conditions at the gate. Assurance independently tests both planes and the gate. Its findings go to the responsible owner for correction.
Three coupled loops
Production creates work and revisions. Assurance challenges and tests. Learning turns verified correction into a change in the next relevant case.
The square's four edge marks represent separate decisions; they are not four green approval lights. This is the book's reference model.
01 / The framework in practice
One case,
four moments.
A customer's payment terms. New evidence changes whether an earlier assessment can support the next decision.
Customer payment terms · An illustrative case from Chapter 9
A claim supports a proposed change.
Assessment v3 supports 60-day terms. No action has been executed.
- EvidenceCustomer accounts
- Claim · v3Liquidity assessment
- Use boundary · openBarrier across the route
- Proposed effect · Net 60 days60-day payment terms
Select a moment to inspect it.
Concept illustration · No action executed.
02 / The argumentThe organization
The organization
behind the intelligence.
An agent produces an answer. Someone copies it into a report. Another system retrieves it. A team acts on it.
Claim, evidence,
Review with the
power to challenge.
Why human judgment and confidence must be assessed alongside machine output.
Read the essaysA mandate at
the moment of use.
A refund recommendation can be perfectly clear and still lack authority.
Try the action boundaryStart here
Read the manifestoRead the ten principlesUse the Four Questions worksheet03 / The open frameworkAn open foundation.
An open foundation.
A shared vocabulary.
The open white paper, version 1.0, defines the model and its limits.
Read on Zenodo
04 / The book
AI-nterprise
Governing Knowledge, Judgment, and Authority in the Age of AI Agents
For the leader accountable for what an organization does with what it believes.
Forthcoming · Q4 2026 · English edition
A Manifesto for the Human-Agent Enterprise
The public argument for governing what humans and agents may know, decide, remember, and do together.
Read the essay
The Model Is Rented. The Learning Must Be Yours.
Why corrections and evaluations must become reusable organizational learning.
Read the essay
The Human Hallucination: The AI Failure Mode Nobody Is Measuring
Why judging the human–AI pair matters when fluent output changes a person's confidence.
Read the essay06 / Practical resourcesTake the questions
Take the questions
into your work.
Four Questions is the complete printable worksheet currently available for the reader.
One open download. No account or signup.
Explore the resourcesPreview worksheets · Version and limitations shown with each file.