We use a few strictly-necessary cookies, no marketing trackers, and cookieless analytics. Read our cookie policy, or pick a preference.
Working papers and longer research on compliance methodology, conformity assessment, and the practice of governing AI systems. The evidence base beneath the platform, published in the open.
Why the exposure from ungoverned AI now runs into the trillions, and why the gap opens precisely where there is no system of record. Anchored on the Allianz Risk Barometer and the IBM Cost of a Data Breach.
The method behind the corpus: decomposing a regime into atomic obligations and archetypes so a single control can be traced to every rule that demands it.
↗Reframing conformity assessment as the maintenance of a verifiable record, and what that changes for how high-risk systems are certified and kept certified.
↗The Triage, Rate, Architect, Control, Evidence model for governing a system that answers to several regimes at once, and how the tiers compose.
↗The gap between the system that was certified and the one in production, and a cryptographic approach to closing it as a tamper-evident chain of custody.