The Answerability Index · Fortune 100 (pilot)Illustrative prototype · placeholder data

We asked five AI systems who leads enterprise cloud infrastructure. They rarely agreed.

A buyer asking five AI assistants the same question meets five different shortlists. The same companies are surfaced at wildly different rates — and some are not surfaced at all.

Observed surfacing & cross-engine divergenceCapture window: ILLUSTRATIVE · n=placeholder
Surfacing rate0%100%· share of prompts in which the company was surfaced
σ — cross-engine divergence (std. dev. across engines); higher = the engines disagree more about this company
How to read this: observed surfacing · within this prompt battery · during this capture window · most-consistently-surfaced, not "best." This measures AI-mediated perception under bounded conditions — not company quality, and not an endorsement.

PROTOTYPE — fictional companies, placeholder numbers, to lock the visual before real capture. The published edition uses real Fortune 100 companies and real engine captures (≥3 runs/engine, dated window). The Answerability Index · prototype 2026-05-26.