Source Authority
Source authority is an AI engine's implicit trust in a domain when selecting which pages to retrieve and cite for an answer. It is shaped by content consistency, entity clarity, structured data, and third-party corroboration rather than backlinks alone — and different engines weight sources differently, so authority must be assessed per engine.
Search-era authority was a link-graph property; answer-era authority is a selection property: does the engine's retrieval pipeline pick your page, and does the model treat what it says as reliable enough to repeat? Pages win selection by being topically exact, extractable, current, and corroborated by other sources the engine also trusts.
Corroboration deserves emphasis. Models are pattern-averagers: a fact stated only on your site competes with every third-party page that says otherwise. When your pricing page, your G2 profile, review articles, and community threads all state the same facts, the engine's synthesized answer converges on them; when they conflict, the answer becomes a coin flip.
Because each engine has its own retrieval stack, authority is not portable. The review aggregator Perplexity leans on may barely appear in Gemini's citations. Per-engine citation data is how you learn who each engine actually trusts in your category.
Related terms
See what AI currently says about your brand — free snapshot.