The Trust Plane

Name: The Trust Plane
Author: Ben Faircloth

Governing AI Behavior Through Evidence, Outcomes, and Earned Autonomy

AI systems are no longer just answering questions. They are retrieving knowledge, writing code, invoking tools, routing work, and shaping decisions.

The Trust Plane gives organizations an operating model for deciding what those systems are allowed to do, how they earn more autonomy, and how governance keeps pace as behavior becomes more consequential.

Join the early reader list

Contents

Twenty chapters. One operating model.

When Answers Become Behavior
Trust Is Not the Starting Point
Evidence Is Not Truth
The Ontology Is Not the Operating Model
Evals Are Not Governance
Human in the Loop Is Not an Operating Model
The Tool Call Is the Boundary
Agents Are Authority Problems
Autonomy Is Earned
What the Trust Plane Is
Behavior Is the Object of Governance
Governance Turns Evidence into Decisions
Contestation Keeps Governance Honest
Intervention Changes Future Behavior
Outcomes Become Evidence
Runtime, Post-Event, Batch, and Fleet Governance
The Minimum Viable Trust Plane
Maturity Is Not More Automation
A Trust Plane in Practice
Governed Autonomy
Endnotes
Appendix: Related Traditions
Appendix: The Trust Plane in Action

The old AI governance question was: do we trust the model?

The better question is: what is this system allowed to do?

As AI moves from chat interfaces into workflows, tools, memory, retrieval, and autonomous action, trust can no longer be treated as a feeling, score, policy, or one-time approval.

Trust becomes operational permission.

The Trust Plane introduces a practical operating model for governing AI behavior through evidence, contestation, intervention, outcomes, and earned autonomy.

Join the early reader list.

Get updates on the manuscript, publication timeline, and early access.