Scorecard · AI Power Chain framework
AI Power Chain risk scorecard
Screen an AI infrastructure target across thirty-nine diagnostic questions covering the six engineering layers between grid interconnection and the accelerator die. Six questions (one per layer) carry a critical DD flag with higher weight. Answer what you know. The tool computes an overall risk score, a per-layer breakdown, the five most material exposures, and a diligence-gap ratio.
How to use this. Pick the option per question that best matches the target's current state. Where you do not know the answer, pick the "not currently tracked" option so the tool can separate diligence exposure from confirmed risk. Standard questions score 0-4 by observed risk. Six critical DD flag questions (one per layer) score 0-5 to reflect their heavier weight in a real DD memo. Selected questions include a risk discriminator explaining what separates a clean answer from a red-flag answer. Full methodology at /framework-ai-power-chain/. Answers save to your browser as you go, so you can leave and resume.
0 answered