Context
Kakimov and colleagues published a peer-reviewed observational framework for auditing citations in Google AI Overviews using YMYL queries drawn from MS MARCO. Their publisher abstract reports that documents classified as AI-generated were cited more frequently than human-authored documents after controlling for observed retrieval rank, particularly among citations outside the observed retrieved results.
What changed
The study formalizes query-document auditing with rank- and provenance-conditioned citation measures, and reports a citation-provenance association in Google AI Overviews that is not fully explained by observed organic retrieval rank.
Why it matters
AI discoverability measurement should separate observed retrieval position from citation selection and record document provenance. Organic rank alone cannot be treated as an AI citation metric.
Visibility OS interpretation
Visibility OS interpretation: stage-specific measurement is preferable to assuming that organic search rank predicts citation selection. The observed association does not demonstrate that creating AI-generated content improves visibility or establish Google's internal retrieval mechanism.
What remains uncertain
How robust are the provenance labels, statistical controls and effect size in the full manuscript? Does the association persist in later AI Overview deployments or other topics? What is the exact internal candidate pool beyond observed organic search results?