Agent assurance is the discipline of establishing justified confidence in work performed by AI agents, through controls, audit trails, standards, and accountable human ownership, on behalf of clients who cannot verify that work themselves.
A book, serialized one chapter each Wednesday from 8 July 2026.
Contents
Part I The Problem Agent Assurance Solves
What changes when AI does real work, and why nobody can simply check it.
- 1The Capability Overhang8 July 2026
- 2The Verification Gap15 July 2026
- 3Controls as Recorded Incidents22 July 2026
Part II Precedent
How assurance formed around risky capabilities before, why the work lasts, and who buys it.
- 4Prior Assurance Regimes29 July 2026
- 5The Economics of the Control Layer5 August 2026
- 6The Buyer and the Engagement12 August 2026
Part III The Practice of Agent Assurance
The controls: close what can be closed, manage the rest by evidence, keep both alive.
- 7The Adoption Doctrine19 August 2026
- 8Closure Controlsscheduled · 26 August 2026
- 9The Evidentiary Layer and the Gatescheduled · 2 September 2026
- 10Continuity, Exit, and the Single Practitionerscheduled · 9 September 2026
Part IV The Profession
What the public record shows, and what it would take for the work to become a discipline.
- 11The State of the Recordscheduled · 16 September 2026
- 12The Formation of the Disciplinescheduled · 23 September 2026