Every other page here measures whether AI mentions you. This one measures whether what it says is correct. A confident, detailed, wrong answer about your products or policies costs you more than not being mentioned at all, and nothing else on this platform will catch it.
These are your branded prompts only. Mark each answer accurate, partly wrong, or wrong. The score is the share of reviewed answers that were accurate, so it stays empty until you have checked something rather than showing a flattering or alarming zero.
In these answers, AI sometimes gives a reason to think twice about you. This list groups those reasons by meaning, counts the answers that raise each one, quotes them, and names the pages those answers cited. It reads answers to questions that name you and questions that do not, and says how many came from each.
Each claim AI made about you, with its wording variants grouped, the approved fact and its source beside what the models said, and who owns the next step. A claim is reviewed wording by wording; grouping never decides it, and every grouping can be undone.
The facts about your business that a person has checked. When you review a claim AI made, the matching fact is shown beside it as the reference. Approving a fact changes no score, and it does not change what an AI model knows.