AI Usage Disclosure

Where AI is used

In some reports, AI helps write narrative explanations — turning underlying evidence into readable prose for a specific name and combination.

What AI never controls

AI does not calculate your Matrix score or verdict. Those come from deterministic, rule-based logic applied to the same evidence visible in your report — see Methodology. AI narrative output is generated after the score is already determined, never the other way around.

Validation and fallback

AI-generated narrative text is checked before it's shown to you. If generation fails, times out, or doesn't pass our validation checks, the report falls back to a plain, deterministic template instead of showing an unvalidated or partially-broken AI response.

What we don't do

We don't use your name inputs to train a public AI model. We don't disclose which specific third-party AI provider or model version is used — that's an internal implementation detail, not something that changes what you should trust or not trust about the output. What matters is the same regardless of provider: AI never sets your score, and unvalidated AI output never ships.