AI Disclosure

We sell measurement of AI systems. A company that measures how AI describes other businesses should be plain about how it uses AI itself. This is the same standard we apply to our own work, not a caveat attached to it.

The AI Recommendation Assessment puts standardized questions to independent AI systems — currently ChatGPT, Claude, Gemini and Perplexity — and records exactly what each one answers. Those answers are the evidence. We do not paraphrase them into something more flattering, and we report disagreement between systems as readily as agreement.

AI-assisted tools help us organize evidence and prepare documents. They do not decide what a briefing says. A person reviews every deliverable before it is issued, and responsibility for what it says rests with us, not with a model.

The free check and the paid report are produced by a deterministic engine that reads your page and applies fixed rules. The same page read twice produces the same result. No language model writes your findings.

Sara answers common questions on this site. She introduces herself as what she is and never presents herself as a person.

Today Sara has no generative capability at all. She draws from a fixed set of reviewed answers about what we do, so she cannot improvise, and cannot invent a claim, a price or a delivery time. When a question falls outside that set she says she does not have enough information and routes it to a person, rather than guessing. If that ever changes, this page changes with it.

A measurement describes what was true at the time of measurement. It is a dated snapshot, not a forecast. AI systems change, and a later measurement that differs is information rather than a contradiction.

Responsible AI · Privacy · Terms of Use