Resources

Practitioner thinking on AI security.

Working guidance from the field — not marketing collateral. Whitepapers on securing AI, building it privately, and getting it audit-ready. Written for the people who actually have to make AI defensible: security leads, engineering leads, and the founders carrying both jobs at once.

Whitepapers

Download the thinking.

Each whitepaper is a short, gated download — a name and work email, and it is yours. No list, no follow-up sequence.

Available now

AI Red Teaming: A Field Guide

Edition 1 · May 2026 · 12 pages

How adversarial testing finds the failures in AI systems before an attacker does. The threat surface mapped to the OWASP Top 10 for LLM Applications, how a structured engagement runs, and how to move from a one-off test to continuous assurance.

Download the whitepaper →
Available now

AI Security & ISO 42001: A Field Guide

Edition 1 · May 2026 · 13 pages

What the international standard for AI management asks of you, where AI systems most often fall short of it, and how to get to a certificate — with ISO 42001 set alongside SOC 2, the NIST AI RMF, and the EU AI Act.

Download the whitepaper →
In progress

More on the way.

Two further pieces in development. Enter your email on any download and we will let you know when each is published.

Coming soon

The Sovereign AI Factory — a reference architecture

How a private, secure AI product is built end to end — the flagship piece tied to the Factory practice.

Coming soon

A SOC 2 & ISO 42001 AI readiness checklist

A practical, one-page-style checklist for getting an AI system audit-ready against SOC 2 and ISO 42001.

Get started

Reading is a fine place to start. Talking is faster.

If a whitepaper raised a question about your own systems, a security review will answer it directly — thirty minutes, no pitch.

Book a security review →