Governance

Standards alignment

We align our evaluation method to recognised AI risk and quality frameworks, so findings map to the language your governance teams already use.

Precisely worded, on purpose. Our method is aligned with and informed by these frameworks. We do not assess conformity against them, and we are not an accredited assessor for any of them.

NIST AI Risk Management Framework

Used to structure risk across the lifecycle — govern, map, measure, manage — so findings sit in a shape risk teams recognise.

ISO/IEC 42001

The AI management-system standard. We borrow its vocabulary so our findings slot into an organisation's existing AI governance rather than sitting in a silo.

OWASP Top 10 for LLM Applications

Informs our adversarial red-team coverage — prompt injection, insecure output handling, excessive agency, sensitive-data disclosure and more.

Recognised safety benchmarks

Public risk and reliability benchmarks inform how we design our safety and robustness tests, alongside the domain-specific cases we build for you.

Framework names belong to their respective owners. Referencing them describes how our method is structured; it does not imply endorsement by, or affiliation with, any standards body.