Three different questions: Position asks how the current score compares with the displayed panel average. Movement asks how much the aggregate score changed during the selected window. Claim drift asks whether stored model comparisons changed materially, even when the aggregate score stayed the same. High/Critical rate is a separate severity axis. These measures should not be treated as interchangeable.

Why this matters: AI visibility tells you whether a company appears. SemanticRisk examines what was understood, how that interpretation compares, and whether the underlying claims remained stable. Open a company record to inspect the source-to-claim evidence chain.