AI brand monitoring: What the assistants believe
about you, checked against
what is actually true.

No, and nothing else can either. There is no interface that edits what an assistant says about a brand. An answer is rebuilt from what the assistant reads, so the only lever anyone has is the content it reads. Verand finds the wrong claim, captures the right one, and gets it stated plainly in pages you own.

You. The correction box is typed by a person who knows the answer, and Verand never decides which claim is wrong. What it does with your answer is fixed: it saves it to your Brand Hub knowledge base as a first-hand fact, embedded on save, so later drafts can retrieve it by topic.

Four concrete things, and nothing beyond them. The correct value is stored as canonical authority in your knowledge base, where later drafts retrieve it by topic. Every correction still outstanding rides into the article editor, flagged relevant where the draft actually discusses the subject, with a one-click weave that previews the edit, re-runs the checks and declines when there is no natural spot for it. Every fact shows the pages the assistants were reading when they stated it and what they said, and on a fact marked Inaccurate each page offers a next step: Fix it on your page, which opens it in Page Tracker, or Ask them to correct it, on someone else's. That drafts a short correction request from your own correction text and nothing else. It never says their page is wrong, because Verand has read the answer that cited the page, not the page; it never mentions AI or rankings; and it runs through your compliance packs before you see it. Verand never sends it: you edit it and send it yourself, or don't. And the page goes live only when a person publishes it. Nothing is sent to an assistant, because there is nowhere to send it.

Each corrected claim carries a status. One that is still being stated reads Still present after every scan. It turns Fixed only when an assistant states your corrected value, which is checked on the same run that re-extracts the claims. A claim that merely stops appearing is left where it is, because extraction rewords claims between runs and absence is not evidence. Weeks, not days.

Then there are no claims to check and the card says exactly that instead of inventing some. Brand Facts only reads answers that name you. Getting named comes first, which is what the Sources and Topics tabs are for.

No. Brand Facts is site-level and lives on the Sentiment tab. Per-article AI-visibility analysis is on the roadmap, is not built, and is not sold. The one per-article surface today is the corrections card in the editor, which carries your outstanding corrections into the draft you have open.