Publications
Where the work is argued in full. The problem, the frameworks, and the architecture that turns obligation into evidence.
Publication Three
The FCA's Next Examiner Is an Agent.
Recommendation 6 of the Mills Review describes supervisory agents testing firm evidence against expectations. The perimeter most coverage misses, and the clock.
Read Publication TwoYour AI Agent Works. Now Prove It.
82 percent of executives trust their AI controls. 14.4 percent of firms approve every agent before production. The four-part architecture that closes the gap.
Read Publication OneThe Regulators Have Stopped Waiting.
COSO, four UK regulators and the EU AI Act are telling the same story. Why the challenger with the better product still loses the deal on evidence.
ReadNew publications, sent when they are published. Nothing else.