The ARC framework systematically evaluates AI brand visibility across three core pillars: Association, Relevance, and Citations. It deploys directional probes to map semantic connections between entities, queries, and brand names, tests models directly to quantify category relevance scores, and mines grounded search outputs to track which URLs and competitor mentions are actively cited in AI-generated answers.