Readiness quiz

How do we answer and calculate the result?

Frontier Development research deskSection 08 · Page 2 of 20Last updated 26 October 2026

Choose one answer for each of 12 questions: A = 0, B = 1, C = 2 points. Total possible score is 24. Unknown is A; partial evidence is B. Do not infer readiness from company size or an impressive demo.

Hard blockers: an A on question 3 (rights), 5 (protected test), 10 (accountable owner) or 12 (risk/processing review) means Blocked for commissioning, regardless of total. A B on any of those questions means Conditional until the gap is resolved. For an external data transfer or operational pilot, all four must reach C and any project-specific legal/security approval must be complete.

TRACE THE EVIDENCE

Section source library

This page contains no direct external citations. These are the section’s references; they do not independently validate the desk’s proposed scoring or methods.

Demystifying evals for AI agents — Anthropic, 9 January 2026 →www.anthropic.com
Llama 3.3 Community License Agreement — Meta →github.com
AI Risk Management Framework — NIST →www.nist.gov
RouteLLM: Learning to Route LLMs with Preference Data — Ong et al., revision 23 February 2025 →arxiv.org
A Field Guide to Rapidly Improving AI Products — Hamel Husain →hamel.dev
GroupKFold — scikit-learn documentation →scikit-learn.org
Pricing — Together AI, snapshot checked 26 October 2026 →www.together.ai
The Inference Engineering Masterclass — Philip Kiely and Ali Taha, Latent Space, 3 August 2026 →latent.space
The Open Source AI Definition 1.0 — Open Source Initiative →opensource.org
SB26-189 Automated Decision-Making Technology — Colorado General Assembly →leg.colorado.gov
Guidance on HIPAA & Cloud Computing — HHS Office for Civil Rights →www.hhs.gov