Perplexity
-
How AI answer engines choose their sources: what’s documented, what’s observed, what’s unknown
What AI companies document about search and citations, what bounded studies observed, and what remains unknown about how sources are selected.
-
Preprint: AI search APIs favored popular brands in Italy; fact-preserving rewrites barely helped
A Politecnico di Milano preprint tested OpenAI, Google and Perplexity APIs on Italian commercial questions. Popular brands got more mentions; fact-preserving rewrites barely moved rank with search off.
-
Retrieval, not FAQ blocks or schema, tracked AI citations in a 158,099-answer preprint
In a University of Toronto preprint, common generative engine optimization (GEO) page tactics showed no statistically significant association with citations once pages were compared within the same website. In a replay through OpenAI’s API, what the model’s own searches retrieved explained far more.